Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Consensus Paper: The Role of the Cerebellum in Perceptual Processes

Consensus Paper: The Role of the Cerebellum in Perceptual Processes Cerebellum (2015) 14:197–220 DOI 10.1007/s12311-014-0627-7 CONSENSUS PAPER Consensus Paper: The Role of the Cerebellum in Perceptual Processes Oliver Baumann & Ronald J. Borra & James M. Bower & Kathleen E. Cullen & Christophe Habas & Richard B. Ivry & Maria Leggio & Jason B. Mattingley & Marco Molinari & Eric A. Moulton & Michael G. Paulin & Marina A. Pavlova & Jeremy D. Schmahmann & Arseny A. Sokolov Published online: 6 December 2014 The Author(s) 2014. This article is published with open access at Springerlink.com Abstract Various lines of evidence accumulated over the past predictive processing, and perceptual sequencing. While no 30 years indicate that the cerebellum, long recognized as single explanation has yet emerged concerning the role of the essential for motor control, also has considerable influence cerebellum in perceptual processes, this consensus paper sum- on perceptual processes. In this paper, we bring together marizes the impressive empirical evidence on this problem experts from psychology and neuroscience, with the aim of and highlights diversities as well as commonalities between providing a succinct but comprehensive overview of key existing hypotheses. In addition to work with healthy individ- findings related to the involvement of the cerebellum in sen- uals and patients with cerebellar disorders, it is also apparent sory perception. The contributions cover such topics as ana- that several neurological conditions in which perceptual dis- tomical and functional connectivity, evolutionary and com- turbances occur, including autism and schizophrenia, are as- parative perspectives, visual and auditory processing, biolog- sociated with cerebellar pathology. A better understanding of ical motion perception, nociception, self-motion, timing, the involvement of the cerebellum in perceptual processes will O. Baumann (*) J. B. Mattingley M. Leggio Queensland Brain Institute, The University of Queensland, St. Lucia, Department of Psychology, Sapienza University of Rome, Rome, Queensland, Australia Italy e-mail: o.baumann@uq.edu.au M. Leggio M. Molinari R. J. Borra I.R.C.C.S. Santa Lucia Foundation, Rome, Italy Department of Radiology and Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard E. A. Moulton Medical School, Charlestown, MA, USA Pain/Analgesia Imaging Neuroscience (P.A.I.N.) Group, Department of Anesthesia, Boston Children’s Hospital, Center for Pain and the R. J. Borra Brain, Harvard Medical School, Waltham, MA, USA Department of Diagnostic Radiology, Medical Imaging Centre of Southwest Finland, Turku University Hospital, Turku, Finland M. G. Paulin Department of Zoology, University of Otago, Otago, New Zealand J. M. Bower Numedon Inc., Pasadena, CA, USA M. A. Pavlova Department of Biomedical Magnetic Resonance, Medical School, K. E. Cullen Eberhard Karls University of Tübingen, Tübingen, Germany Department of Physiology, McGill University Montreal, Montreal, Canada J. D. Schmahmann Ataxia Unit, Cognitive Behavioral Neurology Unit, Laboratory for C. Habas Neuroanatomy and Cerebellar Neurobiology Department of Service de NeuroImagerie, CHNO des Quinze-Vingts, UPMC Paris Neurology, Massachusetts General Hospital and Harvard Medical 6, Paris, France School, Boston, MA, USA R. B. Ivry A. A. Sokolov Department of Psychology, University of California, Berkeley, CA, Département des Neurosciences Cliniques, Centre Hospitalier USA Universitaire Vaudois (CHUV), Lausanne, Switzerland 198 Cerebellum (2015) 14:197–220 thus likely be important for identifying and treating perceptual cerebellum in pain perception is reviewed by Drs. Borra and deficits that may at present go unnoticed and untreated. This Moulton. Dr. Ivry presents a hypothesis and data to suggest paper provides a useful framework for further debate and that the cerebellum acts as a timing device for motor and non- empirical investigations into the influence of the cerebellum motor processes. Drs. Leggio and Molinari present evidence on sensory perception. for a model that posits a central role for the cerebellum in the detection and prediction of perceptual sequences. The review . . . Keywords Audition Biological motion Cerebellum closes with a contribution from Dr. Bower, who suggests that . . . . . Connectivity Evolution fMRI Pain Perception the cerebellum is not itself involved in perceptual processing, . . . . Prediction Single-unitrecording Self-motion Sequencing but instead, its influence on perception as well as motor . . State estimation Timing Vision control, is indirect through its role in monitoring and adjusting the acquisition of sensory data. Introduction Anatomical Circuits Relevant to the Role For 150 years, functional analyses of the cerebellum have of the Cerebellum in Perception (J.D. Schmahmann) focused on the role of this subcortical structure in the control and coordination of movement. In the past 30 years, however, The cerebellar role in perception is predicated on the fact that clinical, experimental, and neuroimaging studies have provided it is an essential node in the distributed neural circuits compelling evidence for the involvement of the cerebellum in subserving sensorimotor, autonomic, and cognitive function task domains as diverse as memory, language, and emotion. as well as emotional processing. The following is a summary Crucially, several lines of evidence suggest that the cerebellum of these pathways and connections. For earlier comprehensive has an influence on perceptual functions. Observations from reviews and original citations, please see Schmahmann [1–3] anatomical and electrophysiological studies in monkeys and and Schmahmann and Pandya [4]. cats indicate the existence of cerebellar connections with visual- and auditory-related cortices. Moreover, clinical reports Peripheral Afferents in humans have revealed that both focal and diffuse lesions of the cerebellum lead to a wide range of sensory impairments. Auditory and visual inputs are conveyed from primary sensory While damage to the cerebellum does not cause a complete loss receptors to vermal lobules VI and VII [5], and visual inputs of sensory function, it is apparent that several sensory and also reach the dorsal paraflocculus. Spinocerebellar tracts ter- perceptual processes are affected, such as motion and time minate in the sensorimotor cerebellum in the anterior lobe and perception, or the ability to recognize perceptual sequences. lobule VIII [6], while vestibular afferents target lobule X [7]. In this consensus paper, we summarize key findings and Climbing fibers from the sensorimotor-recipient inferior olivary concepts with the aim of demonstrating and explaining the nuclei project to the sensorimotor cerebellum; the principal cerebellar influence on perceptual tasks. To this end, we have olivary nucleus is devoid of peripheral inputs and is linked with gathered contributions from 14 experts in complementary the cognitive cerebellum in the posterior lobe (see [3]). fields of psychology and neuroscience, providing a range of different and sometimes controversial viewpoints. We believe Cerebrocerebellar Pathways that a new consensus that draws on and integrates the ideas presented here will help unravel the enigmatic role or influ- Cerebellar connections with the cerebral cortex include two- ence of the cerebellum in perceptual processing. The review stage feedforward and feedback loops with obligatory synap- begins with a succinct overview of the anatomical connections ses in the pons and thalamus. The top-down circuit is of the cerebellum with sensory and perceptual areas in the corticopontine–pontocerebellar and the bottom-up is cerebrum by Dr. Schmahmann. Dr. Habas then provides an cerebellothalamic–thalamocortical. evaluation of the functional connections between the cerebel- lum and cerebral perceptual systems, drawing on studies using Corticopontine Projections modern neuroimaging techniques. Dr. Paulin provides an evolutionary and comparative perspective on cerebellar in- Knowledge of the corticopontine projections provides critical volvement in perceptual functions. Evidence for a cerebellar insights into the nature of the information to which the cere- role in visual and auditory processing is summarized by Drs. bellum has access. Projections arise from neurons in layer Vb Baumann and Mattingley, followed by a commentary from of sensorimotor regions in the precentral, premotor, and sup- Drs. Pavlova and Sokolov on visual processing of biological plementary motor area, primary somatosensory cortices, and motion. Dr. Cullen writes on the critical function of the cere- the rostral parietal lobe [8–11]. Studies in stroke patients also bellum in self-motion perception. Evidence for a role of the show topography of motor function in the human pons [12]. Cerebellum (2015) 14:197–220 199 Considerable corticopontine projections are derived also behavioral manifestations. The superior parietal lobule con- from the prefrontal cortex, multimodal regions of the posterior cerned with multiple joint position sense, touch, and proprio- parietal and temporal lobes, paralimbic cortices in the cingu- ceptive impulses projects throughout central and lateral re- late and posterior parahippocampal gyrus, and visual associ- gions of the rostrocaudal pons. The caudal inferior parietal ation cortices in the parastriate region, supporting multimodal, lobule implicated in the neglect syndrome favors the rostral supramodal, and limbic related functions necessary for per- half of the pons in the lateral and dorsolateral regions [10]. ception (Fig. 1). Auditory association areas in the superior temporal gyrus Prefrontopontine projections arise from dorsolateral and and supratemporal plane are connected with the lateral and dorsomedial convexities concerned with attention and conju- dorsolateral pontine nuclei. Cortices in the upper bank of the gate eye movements (area 8), spatial attributes of memory and superior temporal sulcus activated during face recognition working memory (area 9/46d), planning, foresight, and judg- tasks project to the lateral, dorsolateral, and extreme dorsolat- ment (area 10), motivational behavior and decision-making eral pontine nuclei [14]. Motion-sensitive temporal lobe areas capabilities (areas 9 and 32), and from areas 44 and 45 MT (middle temporal), FST (fundus of the superior temporal homologous to language areas in human [13]. sulcus), and MST (medial superior temporal) also have pon- Posterior parietal association cortices are critical for direct- tine connections [15], but inferotemporal cortex including the ed attention, visual–spatial analysis, and vigilance in the con- rostral lower bank of the superior temporal sulcus, which is tralateral hemispace; lesions are associated with complex relevant for feature discrimination, has no pontine efferents. Fig. 1 Composite color-coded summary diagram illustrating the I through IX are taken. c Patterns of termination within the nuclei of the distribution within the basis pontis of rhesus monkey of projections basis pontis. Other cerebral areas known to project to the pons are derived from association and paralimbic cortices in the prefrontal depicted in white. Cortical areas with no pontine projections are shown (purple), posterior parietal (blue), superior temporal (red), parastriate, in yellow (from anterograde and retrograde studies) or gray (from and parahippocampal regions (orange), and from motor, premotor and retrograde studies). Dashed lines in the hemisphere diagrams represent supplementary motor areas (green). a Medial, lateral, and orbital views of sulcal cortices. Dashed lines in the pons diagrams represent pontine the cerebral hemisphere from which the projections are derived. b Plane nuclei; solid lines depict corticofugal fibers (from [1] and [13]) of section through the pons from which the rostrocaudal levels of the pons 200 Cerebellum (2015) 14:197–220 Thus, the dorsal visual (where) stream concerned with motion and converging pontocerebellar projections led to the sugges- analysis and visual–spatial attributes of motion participates in tion that information from one cerebral cortical area is distrib- the cerebrocerebellar interaction, but the ventral visual (what) uted to numerous sites in the cerebellar cortex [27], although stream governing visual object identification does not. trans-synaptic viral tract tracing studies reveal that anterograde Parastriate projections from occipitotemporal and occipitoparietal projections through the medial pons are directed to focal areas regions also respect the dorsal–ventral dichotomy. The medial in crus I and crus II [28]. and dorsal prelunate regions project to the pons (dorsolateral, lateral, and lateral aspect of the peripeduncular nuclei most Cerebellar Feedback heavily), but ventral prelunate cortices and inferotemporal re- gions do not [16]. Projections from the temporal lobe homologue Purkinje cells convey the output of the cerebellar cortex to the of the Wernicke language area in human, together with those deep cerebellar nuclei (DCN), which send projections back to from the monkey homologue of Broca’s area, are relevant in the the brainstem, or to the cerebral cortex via the thalamus. The light of cerebellar activation during functional neuroimaging cerebellar cortex–DCN–thalamus–cerebral cortex feedback loop studies of language [17, 18] and in disorders of language follow- is arranged so that motor related interpositus nuclei (globose and ing cerebellar lesions [19, 20]. emboliform in human) send efferents from cerebellar anterior Paralimbic projections arise from posterior parahippocampal lobe motor areas to the cerebral sensorimotor regions, whereas gyrus important for spatial attributes of memory, directed to the ventral dentate sends information from the cerebellar poste- lateral, dorsolateral, and lateral peripeduncular nuclei. Cingulate rior lobe to cerebral association areas—prefrontal, posterior pa- cortex projections arise from motor areas in the depth of the rietal, and others [28, 29](seeFig. 2). The cerebellar vermis and cingulate sulcus [21] and from areas concerned with motivation fastigial nucleus are linked with brainstem and thalamic struc- and drive in rostral and caudal cingulate areas [22]. The anterior tures concerned not only with vestibular and oculomotor control, insular cortex, important for autonomic systems and pain mod- posture, and equilibrium, but also with autonomic and paralimbic ulation also has pontine connections [9]. Projections arise also cerebral areas, consistent with the notion of the vermis and from multimodal deep layers of the superior colliculus and fastigial nucleus as the limbic cerebellum [3]. medial mammillary bodies involved in memory and emotion [23]. The hypothalamus, critical for autonomic control and lim- Synthesis bic behaviors, has direct reciprocal connections with the cerebel- lum [24]. Against the backdrop of the heterogenous and topographically Corticopontine projections are arranged with topographic arranged connections of the cerebellum with the rest of the specificity. Sensorimotor terminations are more caudally situated; neuraxis stands the essentially constant architecture of the association areas project more rostrally. Terminations occur in cerebellar cortex. This dichotomy is the basis of the dysmetria multiple patches forming interdigitating mosaics. The signifi- of thought theory, which poses that a constant computation— cance of associative corticopontine inputs in human compared the universal cerebellar transform—is applied to multiple with monkey is underscored by enlargement in human of the domains of neurological function subserved by the distributed medial part of the cerebral peduncle conveying prefontopontine neural circuits of which cerebellum is an integral node [3]. The fibers [25], reflecting evolutionary pressure in which intercon- anatomical connections that link the cerebellum with both the nected systems evolve in concert with each other. external and the internal worlds thus provide the critical neural substrates of the putative cerebellar role in perception. These Pontocerebellar Projections conclusions from tract tracing studies in the monkey are supported by resting state functional connectivity magnetic The caudal pons sends sensorimotor-related information to the resonance imaging (MRI; [30]) and task-based functional cerebellar anterior lobe. Rostral pontine nuclei convey cogni- MRI studies in humans [18], as well as by clinical investiga- tively relevant information to the posterior cerebellum: medial tions in patients with cerebellar damage [19]. pontine projections from prefrontal cortices to crus I and to crus II, and medial, ventral, and lateral pons conveying infor- mation from parietal association cortices to crus I, crus II, and lobule VIIB. These anatomical studies extend earlier physio- Resting-State Functional Connectivity logical conclusions that parietal and prefrontal cortices are Between Cerebellum and Sensory Systems (C. Habas) functionally related mainly to crusI,crusII, andthe paramedian lobule of the cerebellum [26]. In the pontocerebellar projection, Measurement of human brain resting-state activity with MRI each cerebellar folium receives input from a unique comple- has allowed us to precisely determine the functional connec- ment of pontine cell groups, some of which are widely separat- tivity (FC) between specific zones of the cerebellum and the ed [1, 27]. The pattern of diverging corticopontine projections rest of the brain. FC is based on temporal correlations between Cerebellum (2015) 14:197–220 201 Fig. 2 a Diagram of the lateral view of a cebus monkey brain (top) to show the location of injections of McIntyre-B strain of herpes simplex virus I in the primary motor cortex arm representation (M1arm), ventral premotor cortex arm representation (PMVarm), and in the prefrontal cortex in areas 9 and 46. The resulting retrogradely labeled neurons (below)in the cerebellar interpositus nucleus (IP) and dentate nucleus (DN) are indicated by solid dots and show the dorsal–ventral dichotomy in dentate projections to motor versus prefrontal cortices. Adapted from [29]. b Representation on flattened views of the cerebellum of the input– output organization of cerebellar loops with motor cortex M1 (left) and area 46 (right) revealed using anterograde and retrograde strains of rabies virus as tract tracer. M1 is interconnected with lobules IV to VI; prefrontal cortical area 46 is linked predominantly with crus II. Adapted from [28] spontaneous, low-frequency (0.01–0.1 Hz) fluctuations of the lobules IV–VI, the second in lobules VIIb–VIII, and a third blood-oxygen-level-dependent (BOLD) signal at rest between in lobules VI–VIIA [36]. functionally and anatomically linked cerebral areas [31]. Two Discrepant results, however, were obtained for the visual main statistical methods are used to compute resting-state and auditory cerebellum. O’Reilly and colleagues [34]found functional maps passing through the cerebellum [32]: (1) functional coherence between visual area MT and superior independent component analysis, which is used to identify temporal gyrus, including auditory primary and associative multiple temporally cohesive, spatially distributed networks zones, with cerebellar lobules V-VI-VIII and lobules V–VI, and (2) regression analysis of activity in a region of interest respectively. Buckner and colleagues, however, failed to de- against that of the remainder of the brain. These methods tect any functional connectivity between auditory cortex and have contributed to distinguish two anatomo-functional cerebellum [30, 33]. The proximity between the occipital lobe parts of the cerebellum [33–35]: a sensorimotor region and the underlying cerebellar cortex has been proposed as a (lobules V–VI and VIII) and a prominent multimodal cog- possible explanation of the discrepancy between these data. nitive and limbic region (lobule VIIA, especially crus I and However, Sokolov et al. [37] (see also the section by Drs. II, with adjoining parts of lobule VI and VIIB, and lobule Sokolov and Pavlova, “Cerebellar Involvement in Biological IX). The sensorimotor cerebellum corresponds predomi- Motion Processing (A.A. Sokolov and M.A. Pavlova)”) nantly to sensory parts of its multiple somatotopic maps found, using DTI, structural interconnection between cerebel- that receive exteroceptive and proprioceptive inputs from lar crus I and right superior temporal sulcus (STS), in agree- spinal, trigeminal, and somatosensory cortical afferents, and ment with a previous seed-based functional connectivity result send outputs to motor areas in order to control, guide, and which showed functional coherence between STS and cere- correct ongoing movements. At least three somatotopic bellar lobules VI/VIIA [38]. It is noteworthy that no functional representations have been reliably described: the first in link was found in these two studies between cerebellum and 202 Cerebellum (2015) 14:197–220 primary visual cortex (BA 17) in line with previous animal these animals, the cerebellum is evidently involved in tracking studies. Notwithstanding, using cerebellar seed-based function- objects using the electric sense [46–48]. But comparative ana- al connectivity, Sang and colleagues [39] found correlations tomical, physiological, and behavioral evidence indicated that between visual networks and hemispheric lobules I–VI and this is not an anomaly. Across all vertebrates, the cerebellum vermal lobules VIIb–IX, as well as auditory networks and seems to have a primary role in motion analysis and motion hemispheric lobules VI-VIIb-VIIIa. Ding et al. [40] also iden- prediction, with a role in motor control a consequence of this tified decreased functional connectivity between visual cortex perceptual capability, analogous to the role of dynamical state (BA 17) and cerebellum (crus I and II, vermis of lobules VI–VII estimators in artificial control systems [49]. and tonsilae) when they compared ambliopic patients with The theory that cerebellum is a neural analogue of a dy- healthy subjects. One possibility would be that amblyoply first namical state estimator simplifies and generalizes the theory induced diminished connectivity between primary visual cortex that cerebellum is engaged in motor control. An animal needs and interconnected parietal (BA 40) and prefrontal (BA 6/8) to determine the kinematic state of its own body in order to cortices, and that this altered connectivity indirectly affected the control movements, and to perceive and dynamically interact cerebellum via the prefronto-parieto-pontine pathway. with other objects and organisms. In particular, active sensing The cerebellum is also involved in the limbic ‘salience and exploratory behavior is critically dependent on accurate network,’ mainly encompassing insula, frontal operculum, information about the configuration and motion of sense medial prefrontal cortex, and hypothalamus [35], and in organs during sensory acquisition [47]. It has been shown in charge of interoceptive and autonomic processing [41]. There- human and animal studies that the cerebellum plays a crucial fore, it could be hypothesized that cerebellar zones belonging role in active sensory acquisition [50, 51]. Other tasks that to the salience network (lobules VI, VIIA, and VIIB) process have been shown to involve the cerebellum in humans also interoceptive data. Paravermal and vermal lobule VI may seem to require dynamical state estimation [52–59]. constitute a specific node receiving exteroceptive and intero- The cerebellum is a characteristic of vertebrates, but ceph- ceptive data, since it has been found active during emotional alopod molluscs (squid and octopus) appear to have evolved a responses such as disgust [42]. In conclusion, functional con- cerebellum independently. The cephalopod cerebellum re- nectivity mainly confirms previous results acquired with his- ceives visual and vestibular sense data and is involved in tological tracking and electrical stimulation and adds some whole-body and oculomotor stabilization during locomotion new insights: the ‘sensory’ cerebellum is mainly part of the [60–62]. Cephalopods are the only agile predators among sensorimotor (and vestibular) cerebellum and may also com- molluscs. prise areas that process visual, auditory, and interoceptive Cerebellar-like structures occur in a number of animal signals. Finally, there may be two distinct roles for the cere- phyla. These are distinguished from the ‘cerebellum proper’ bellum in perceptual tasks. The first involves the ‘sensory’ by a lack of climbing fibers and a lack of direct projections to cerebellum for perceptual analysis, cancellation, and anticipa- motor and premotor structures. The most well-known cere- tion based on internal models during, for instance, fine ex- bellar-like structures are electrosensory and lateral-line ploratory movements. The second involves the polymodal mechanosensory nuclei in fishes [63, 64], but they are found ‘executive’ cerebellum, which is associated with working in many vertebrates including humans [65]. They are involved memory, attention, and decision-making processes for con- in removing distortions from external signal sources caused scious elaboration of the mental representation of a perceived by an animal’s own activity. Thus, in electroreception, the object [43]. cerebellum is involved in sensing external targets by exploiting distortions in signals generated by the animal’s own activity, while cerebellar-like circuits are involved in Evolutionary Perspectives on Cerebellar Function (M.G. sensing external targets by eliminating distortions of target Paulin) signals caused by the animal’s own activity. Cerebellar-like circuits have been reported among arthro- Early in the twentieth century, studies of brain-damaged sol- pods, onychophorans, and polychaete annelids. These inver- diers led to a consensus that cerebellum is dedicated to motor tebrates are all active foragers, with appendages that support control, because focal cerebellar ablation led to obvious motor arrays of sensilla [66]. Cerebellar-like structures in insects deficits without obvious perceptual deficits [44]. Late in the may be involved in orientation and navigation [67]. They twentieth century, human functional imaging studies revealed seem to be more prominent in species like honeybees, which that the cerebellum is actively engaged in a variety of cogni- use their antennae as active probes, than in moths whose tive, perceptual, and behavioral tasks, even when subjects are antennae are passive receivers [66]. not moving [45]. The cerebellar cortical circuit common to the cerebellum In the middle of the twentieth century, the gigantocerebellum and cerebellar-like circuits has apparently evolved indepen- of weakly electric fish stood out as an anomaly because, in dently in at least five groups of animals: vertebrates, Cerebellum (2015) 14:197–220 203 cephalopod molluscs, arthropods, onychophorans, and poly- predominantly in vermal lobule VII and hemispheric lobules chaete annelids. All species in which cerebellar and/or VI, that were differentially activated for visual stimuli and cerebellar-like circuits have been reported are motile and auditory stimuli. In the 1980s, several laboratories started to sufficiently large that their kinematics is influenced by inertia, use neuronal tracers to examine cerebrocerebellar projections and they interact with other such animals. Inertia constrains in non-human primates and discovered that visual as well as how the kinematic state (position, configuration, and rates of auditory association areas are anatomically connected with the change) of an object changes as a function of applied force, cerebellum [2] (see also the section by Dr. Schmahmann, such that, if an object has inertia, then information about its “Anatomical Circuits Relevant to the Role of the Cerebellum kinematic state can be used to predict its future position and in Perception (J.D. Schmahmann)”). Interestingly, while cer- configuration at least in the short term. This is not true of ebellar connections were found for dorsal visual stream areas, animals (or indeed objects of any kind) whose mass is small or which are known to underlie motion analysis, this was not the drag is large relative to applied forces [68]. Animals that have case for ventral visual stream areas, which are involved in evolved cerebellar(-like) circuits are, therefore, animals for visual object recognition. This finding suggests that the cere- which probabilistic inference about the kinematic states of bellum is particularly involved in processing dynamic (i.e., self and others is both possible and useful. The fact that this time varying) visual information. group includes disparate, unrelated species indicates that the The first evidence in humans for a cerebellar involvement genetic and developmental capacity for cerebellar(-like) cir- in visual processes derives from work undertaken by Ivry and cuits may be shared by all animals with nervous systems and Diener, who found that cerebellar patients were impaired in that it has been co-opted by evolution whenever there has been making judgments of the velocity of moving stimuli, whereas an ecological opportunity for animals capable of dynamic elementary visual functions remained intact [86]. These find- motion prediction and control [69]. More generally, the ability ings were later corroborated and extended by Thier and to predict state trajectories of dynamical systems from obser- Haarmeier, who reported that patients with cerebellar lesions vations provides a core capability that may underpin a wide were also impaired in detecting and discriminating moving variety of perceptual, cognitive, and motor tasks [70]. visual signals in the presence of visual noise [87]. Similarly, it Until a few years ago, the Kalman filter was the only was found that cerebellar lesions can disturb auditory process- known practical algorithm for dynamical state estimation ing, by significantly increasing thresholds in duration [88]and [71]. It assumes linear target dynamics, an assumption that pitch discrimination tasks [57]. does not hold for mechanical linkages like human and animal Despite evidence of a sensory processing role for the cere- bodies. Newer algorithms based on drawing random samples bellum, the exact manner in which visual and auditory informa- from probability distributions defined by observations are able tion is represented in the human cerebellum remains unclear. To to track states of high-dimensional nonlinear systems [72]. address this issue, we used functional magnetic resonance im- These algorithms can be implemented using spiking neurons, aging (fMRI) to monitor neural activity within the cerebellum in which a spike at a particular location in a network represents while participants were engaged in a task that required them to a sample at a particular location in the state space of the system determine the direction of a visual or auditory motion signal in tracked by the network [73, 74]. There is growing evidence noise [89]. In the visual motion task, vermal lobule VI and right- that neurons use Bayesian Monte-Carlo algorithms of this hemispheric lobule X were active (see Fig. 3a), whereas in the kind to implement decisions and actions [75–83]. auditory motion task, activity was elevated in hemispheric lobules VI and VIII (see Fig. 3b). Interestingly, for both auditory and visual motion tasks, activity within left crus I increased as The Role of the Cerebellum in Visual and Auditory the strength of the motion signal decreased (see Fig. 3c), sug- Processing (O. Baumann and J.B. Mattingley) gesting that the recruitment of the cerebellum is related to the perceptual demands of a task. These findings are consistent with Over the last decade, hypotheses of human cerebellar function results from a positron emission tomography study in which have undergone dramatic revisions [84]. Of these, perhaps the similar regions of cerebellar cortex became more active as the most intriguing is the proposal that the cerebellum plays a role level of difficulty of a pitch discrimination task increased [90]. in sensory processes. In the following, we review evidence for In addition, recent neuropsychological and neuroimaging stud- cerebellar involvement in visual and auditory perception. ies have implicated left crus I in tasks involving biological Cerebellar responses to auditory and visual stimulation motion perception [91, 92] (see also section by Drs. Sokolov were described in the 1940s. Snider and Stowell [85]recorded and Pavlova, “Cerebellar Involvement in Biological Motion electrical responses in the cerebellar cortex of 150 anesthe- Processing (A.A. Sokolov and M.A. Pavlova)”), suggesting a tized cats, evoked by acoustic clicks as well low-intensity light role in higher-level visual processing. flashes. Using this approach, they revealed the existence of Interestingly, there have also been incidental reports of cere- distinct, but partially overlapping cerebellar regions, bellar activity during tasks involving crossmodal matching 204 Cerebellum (2015) 14:197–220 Fig. 3 MR brain slices showing distinct set of cerebellar regions that motion condition; green shading represents activity for the auditory were differentially activated for: a visual stimuli and b auditory stimuli, as motion condition; yellow shading indicates activation overlap between well as c showing a negative linear relationship between fMRI signal and the visual and auditory conditions). Figure reproduced with permission motion signal strength (red shading represents activity for the visual from [89] [93–95]. For example, we observed that combined audiovisual Cerebellum and Perception: The Role of the Cerebellum in motion detection led to increased activity bilaterally in cerebellar Self-Motion Perception (K.E. Cullen)”) is essential for a wide lobule VI and right lateral crus I, relative to unimodal visual and range of daily life activities such as safe car driving, motor auditory motion tasks [93]. This is consistent with findings in learning, imitation, social interaction, and non-verbal commu- monkeys that different sensory areas of the cerebral cortex con- nication through body language [99]. Healthy adults and verge on common areas within the neocerebellum [1]. Taken children easily recognize personality traits through actions of together, these results suggest that the cerebellar hemispheres others, even if they are represented through a set of light dots play a role in the detection of intermodal invariant temporal– placed on the main body joints, in “point-light biological spatial features in concurrent streams of audio-visual information. motion displays” [100, 101](see Fig. 4a). Neurophysiological A prominent hypothesis is that the cerebellum aids infor- and lesional research has revealed the core components of the mation processing by making predictions, in the form of an cortical system underlying visual perception of body motion “internal model” of sensory events [96]. An alternative ac- that includes areas in the frontal [102] and parietal [103–105] count is that the cerebellum facilitates perception by monitor- cortices, the fusiform gyrus and superior temporal sulcus ing and coordinating the acquisition of sensory information (STS) [105–107], mainly in the right brain hemisphere [97] (see the section by Dr. Bower, “Is the Cerebellum Sen- [108]. Yet, our knowledge on engagement of brain structures sory for Motor’s Sake, or Motor for Sensory’s Sake? (J.M. outside the cerebral cortex is still rather limited. Bower)”). A third hypothesis is that the cerebellum functions as Early positron emission tomography (PET) data suggest an internal timing device for both motor and perceptual process- activation of the amygdala and left lateral cerebellum for es, with different regions of the cerebellum thought to provide point-light dance-like biological motion [103]. fMRI also separate timing computations for different tasks [98](seethe indicates cerebellar activity during visual processing of body sectionbyDr. Ivry, “Sensory Processing and the Cerebellum: motion. However, the outcome is controversial, in particular, Timing (R.B. Ivry)”). At present, there is no unequivocal support in respect to topography and lateralization. Right midline for any one of these models, and in fact, each can provide at least cerebellar response was found for a contrast of canonical a partial account for many of the relevant findings. against scrambled point-light actions when observers per- In conclusion, while there is considerable evidence that the formed a one-back repetition task [109]. In a two-alternative cerebellum contributes to auditory and visual sensory process- forced choice (2AFC) discrimination task, bilateral activation es, its precise role is not yet well understood. We need more in the cerebellar hemispheres was shown for canonical and information about how the cerebellum interacts with visual scrambled point-light displays pooled together and contrasted and auditory networks, particularly in terms of the nature against baseline, with specific activation of the left lateral (inhibitory or excitatory) and directionality (feedback or cerebellar region QuP (posterior quadrangular lobule or lobule feedforward) of these connections. VI) when judging direction of biological motion [104]. Psychophysical data in patients with tumors to the left cerebellum showed that damage to the lateral lobules VIIB, Cerebellar Involvement in Biological Motion Processing VIIIA, and crus I and II substantially affects visual sensitivity (A.A. Sokolov and M.A. Pavlova) to biological motion simultaneously camouflaged by addition- al moving dots (a spatially scrambled display containing the Visual perception of bodily movements of others (for percep- same characteristics as a canonical biological motion display tion of self-motion, see the section by Dr Cullen, “The (except for the spatial positions of the dots) served as a control Cerebellum (2015) 14:197–220 205 Fig. 4 Loop between the cerebellum and superior temporal sulcus (STS) Elsevier Inc., with permission of the publisher, Elsevier. c Three- subserving biological motion perception. a Example of a point-light dimensional representation of the structural loop pathway between the biological locomotion stimulus with 11 dots placed on the main joints right STS and crus I, as revealed by diffusion tensor imaging (DTI). of the walking human body. Outline added for illustrative purpose. From Fibers descending from the STS to the cerebellum pass through the pons [246] Pion Ltd., London, www.envplan.com. b Dynamic causal and the middle cerebellar peduncle (MCP), while ascending fibers pass modeling shows reciprocal effective communication between the right through the superior cerebellar peduncle (SCP) and the thalamus. From posterior STS and the left lateral cerebellar lobule crus I during visual [37], copyright © The Author 2012. Published by Oxford University processing of biological motion (BM) that modulates the back connection Press from the cerebellum to the STS. Adapted from [92], Copyright © 2011 for biological motion specificity in this series of studies) [91]. possibility for structural connection between the temporal In contrast, sensitivity was not impaired in patients with cortex and cerebellum had been detected by diffusion tensor lesions to the medial left cerebellum. In accord with lesional imaging (DTI) in non-human primates and humans [25]. By data, fMRI in a homogeneous group of healthy human adults using high-resolution acquisition sequences and optimized indicated activation of the left lateral cerebellar lobules crus I processing, our latest DTI work indicates a bidirectional struc- and VIIB [92]. Convergent lesion and brain imaging findings tural loop between regions in the left cerebellar lobule crus I provide reliable evidence in favor of involvement of the left and right STS that were functionally defined during visual lateral cerebellum in visual processing of human locomotion. processing of biological motion [37] (see Fig. 4c). Moreover, dynamic causal modeling demonstrated bidirec- In neuropsychiatric conditions such as schizophrenia or tional task-related effective connectivity between the left lat- autistic spectrum disorders (ASD), impaired biological motion eral cerebellar lobule crus I and the right STS during body processing [99, 113, 114] and altered cerebro-cerebellar con- motion perception [92](see Fig. 4b). The findings suggest that nectivity [115, 116] represent two major characteristics. Yet the cerebellum interacts with the cortical structure considered the relationship between these characteristics has not been as a hub of the neural network subserving visual processing of experimentally investigated. Reciprocal loops between the biological motion [105–107]. This may account for effects of cerebellum and STS in visual processing of body motion left lateral cerebellar lesions on visual tuning to biological may account for lower STS response to biological motion in motion [91]. children with ASD [117] and help to explain how social While closed cerebellar loops with the frontal and parietal deficits relate to disintegrity of the left superior cerebellar cortices are thought to underlie a variety of cognitive functions peduncle [118] hosting the back connection from the cerebel- [110], direct communication between the temporal cortex and lum to the STS [37]. Cerebellar involvement in biological cerebellum during a visual perceptual task had not been pre- motion processing instigates further research on social brain viously shown. Neuroanatomical evidence in non-human pri- networks in neuropsychiatric conditions. mates points to direct projections from the STS to the pons [8, In a nutshell, the left lateral cerebellum appears to be 9, 14, 111, 112] and from the pons to the cerebellum [27, 108]. strongly involved in visual processing of biological motion However, there has been lacking evidence for a back connec- [91, 92]. This engagement occurs likely through direct recip- tion from the cerebellum to the STS. Resting state fMRI rocal communication with the right STS [37, 92], a keystone analyses (see the section by Dr. Habas, “Resting-State Func- of brain networks for body motion processing and visual tional Connectivity Between Cerebellum and Sensory Sys- social cognition [99, 105–107]. Both specificity of deficits in tems (C. Habas)”) indicated possible functional connectivity patients with cerebellar lesions and network topography in between the cerebellum and temporal cortex [33–35]. A healthy adults suggest that cerebellar engagement in 206 Cerebellum (2015) 14:197–220 biological motion processing and action observation goes Recent electrophysiological analyses of the vestibulo- beyond a general role of the cerebellum in visual motion cerebellum and vestibular sensory pathway of monkeys have processing [86, 119; see also the section by Drs. Baumann provided important insights into the specific neural computa- and Mattingley, “The Role of the Cerebellum in Visual and tions underlying the integration of multimodal information Auditory Processing (O. Baumann and J.B. Mattingley)”]. required for self-motion perception. Recent data indicate a remarkable potential for recovery of First, to generate an accurate perception of our motion visual body motion processing following neurosurgical left relative to the world, the brain must continuously account cerebellar lesion removal and suggest that reorganization in for the omnipresent force of gravity. The brain constructs the cerebellum may trigger topographic shifts in the com- internal models of the world’s physical laws to dissociate tilt municating superior temporal areas [120]. The exact func- from translation by combining inputs from the vestibular tion of the cerebellum within the circuitry for perception of otoliths (which detect linear motion for both movements) with biological motion needs further clarification. Engagement inputs from semicircular canals (which detect rotational mo- of both the left cerebellum and right STS has been reported tion, and thus only respond to tilts) [132]. Consistent with this in emotion recognition through body motion [121], detec- proposal, single nodulus-uvula neurons create an internal tion of social interaction and animacy attribution in Heider- model that accounts for the physics of our world. Notably, and-Simmel movies depicting geometric shapes [122–124], neuronal responses to rotations are modulated as a function of imitation [125], and audiovisual integration ([93]; see sec- head orientation relative to gravity (reviewed in [127]) and tion of Drs. Baumann and Mattingley, “The Role of the different subclasses of Purkinje cells encode head translation Cerebellum in Visual and Auditory Processing (O. versus tilt [133]. This representation of translation could po- Baumann and J.B. Mattingley)”). Effective connectivity tentially be combined with the visual and proprioceptive input between the cerebellum and STS during animacy attribution to provide an estimate of heading direction that is based on has recently been demonstrated [124]. Further studies are information from multiple sensory systems. needed to clarify whether and how communication between Second, to perceive body motion independently of head the cerebellum and STS might underlie other social cogni- motion, the brain must compare vestibular and neck-related tive functions, and to address compensatory potential in inputs. Direct evidence for this computation has been revealed congenital, degenerative, and focal cerebellar affections. in the output of the cerebellum, at the level of the neurons in the most medial of the deep cerebellar nuclei (i.e., fastigial), which comprises two distinct populations of neurons. One neuronal population responds to both externally applied ves- The Cerebellum and Perception: The Role tibular and neck-proprioceptive stimulation, and encodes of the Cerebellum in Self-Motion Perception (K.E. Cullen) body-in-space motion. The other neuronal population only responds to externally applied vestibular inputs and encodes The cognitive representation of self-motion is vital to our head-in-space motion [134]. Notably, the convergence of everyday activities. For instance, walking down a busy city vestibular and proprioceptive inputs in body coding cerebellar street requires an accurate estimate of our own motion relative neurons is non-linear [134] and likely underlies the transfor- to objects in the surrounding complex, three-dimensional mation of vestibular signals from a head to a body reference environment. Self-motion requires the integration of sensory frame in the deep cerebellar nucleus [135, 136]. information from multiple systems including vestibular (head Finally, to ensure perceptual stability in everyday life, our motion), visual (optic flow), proprioceptive, and somatosen- brains must continually distinguish between self-motion that sory (body motion), as well as efference copy motor command is the result of our own (active) movements versus externally signals (reviewed in [126]). applied (passive) motion. Theoretically, the computation of There is strong evidence that the cerebellum, and, in par- passive motion requires a comparison between an internal ticular, the vestibulo-cerebellum, makes vital contributions to estimate of the sensory consequences of active self-motion self-motion perception. First, it has long been known that (i.e., forward model) and the actual sensory feedback lesions of the nodulus and uvula (Larsell’s lobules X and (reviewed in [126]). Cerebellar output neurons dynamically IX) alter the temporal and three-dimensional spatial process- encode this difference during self-motion; fastigial neurons ing of vestibular information (reviewed in [127]). More re- are insensitive to active motion and encode an explicit repre- cently, it has been further shown that visually induced illu- sentation of passively applied self-motion [137]. Specifically, sions of self-motion preferentially activate these same lobules the two distinct fastigial nucleus populations (described in the [128, 129] and that self-motion perception is diminished in paragraph above) selectively and dynamically encode passive patients with midline lesions impacting these regions [130, head and body motion relative to space. Moreover, our evi- 131]. Thus, the vestibulo-cerebellum is thought to be required dence to date suggests that this cerebellar-dependent mecha- for computing the internal representation of self-motion. nism uses an internal model of the expected sensory Cerebellum (2015) 14:197–220 207 The perception of pain itself is a complex subjective experi- ence that incorporates sensory, affective, and cognitive com- ponents. Though neuroimaging studies indicate that the cere- bellum responds to noxious stimuli, its functional relevance in relation to these different dimensions is only starting to gain attention. Ascending Nociceptive Input to the Cerebellum Well-controlled studies of pain often use acute experimental stimuli to activate nociceptive pathways, the physiological processes underlying pain perception. Nociceptors are prima- ry afferents that respond to high threshold mechanical and heat stimuli, and can also respond to chemical stimulation, such as during inflammation. Two major categories of noci- ceptive afferents have been classified: A-delta and C-fiber Fig. 5 The cerebellum integrates sensory input (green boxes) from nociceptors. A-delta nociceptors are thinly myelinated and multiple systems including: (1) the vestibular, (2) visual, (3) fast conducting (>2 m/s), while C-fiber nociceptors are unmy- proprioceptive and somatosensory, as well as from (4) motor efference copy signals. Cerebellar output neurons send ascending projections to the elinated and slower conducting (<2 m/s). Electrophysiological thalamus, hippocampus, and superior colliculus, which in turn connect studies in rodents and cats indicate that stimulation of cutane- the cerebellum to numerous cortical regions (red boxes) that mediate ous and visceral nociceptors, in the form of A-delta and/or C- spatial navigation and voluntary motor control fiber primary afferents, can activate and modulate Purkinje cell activity in the cerebellum [141, 142]. At least two possible nociceptive spinocerebellar pathways have been proposed: (1) consequences of active head motion to selectively cancel a spino-olivocerebellar pathway that conveys A-delta and C- responses to active motion. fiber nociceptive afferent input to Purkinje cells in the cere- In summary, computations in the vestibulo-cerebellum bellar anterior lobe ipsilateral to stimulation [143]and (2)a underlie the transformation of input signals into repre- spino-pontocerebellar pathway conveying C-fiber nociceptive sentations that are essential for self-motion perception input to Purkinje cells in the cerebellar vermis [144]. Details (Fig. 5). Interestingly, these same cerebellar-dependent regarding these putative pathways have been vastly computations likely also contribute to mapping spatial understudied. representation in the hippocampus (Fig. 5, ascending pathwayinred). Notably, ‘place cell’ tuning is impaired Descending Cortical and Subcortical Input to the Cerebellum in mutant mice with cerebellar function deficits [138]. The cerebellum likely shapes the directional tuning of In addition to afferent input, the cerebellum receives input place cells via indirect projections from the deep cere- from brain areas associated with nociceptive processing, in- bellar nuclei. Moreover, ascending projections terminate cluding cognition, affect, and motor function [141]. Our cur- in regions of the thalamus [139]knowntoterminate in rent understanding of the neural basis of pain and its modula- parietal cortex, a region that is vital for spatial naviga- tion includes the somatosensory cortices, periaqueductal gray, tion,aswellasmotor andpremotor cortex[140]. Future anterior cingulate cortex, dorsolateral prefrontal cortices, basal work in monkeys and mice using both passive and ganglia, hippocampus, hypothalamus, and the amygdala active motion are needed to fully understand the impact [145], all of which have connectivity with the cerebellum [1, of the cerebellum on how the hippocampus and cortex 96]. With the cerebellum receiving both descending informa- shape spatial navigation. tion from other brain areas and ascending nociceptive infor- mation from the spinal cord, the structure is ideally positioned to influence, or to be influenced by, the processing of pain. Pain and the Cerebellum (R.J. Borra and E.A. Moulton) Neuroimaging Responses to Pain in the Cerebellum The cerebellum is one of the most consistently responsive brain structures to painful stimuli [141]. While our classical A meta-analysis of 47 neuroimaging studies featuring exper- understanding of this structure suggests that it is involved in imental pain revealed specifically localized responses within the motor response to pain, contemporary thinking indicates the cerebellar vermis and bilaterally in the posterior hemi- spheres [141]. The spatial extent of vermal activation spanned that it may have a more direct role in the processing of pain. 208 Cerebellum (2015) 14:197–220 across vermal lobules III, IV, and V, while the bilateral hemi- studying cerebellar-dependent timing [150, 151]. This form spheric activation spanned from hemispheric lobule VI to crus of learning is only adaptive if the animal is able to represent I. Using the same method of meta-analysis, a similar pattern of the temporal relationship between two sensory events, the activation was observed across 16 neuroimaging studies fea- conditioned and unconditioned stimuli. Importantly, the con- turing pathological pain, in the form of spontaneous pain or ditioned response persists following lesions of the cerebellar aggravation of a clinical condition. cortex but loses its adaptive timing [152](see Fig. 6a). Sen- Though pain neuroimaging studies are not typically de- sory timing as a constraint on motor control is also evident in signed to evaluate the physiological significance of cerebellar many tasks involving volitional movements. To intercept a responses, a few notable studies have focused in on this moving object, the movement has to anticipate the trajectory structure in the context of pain. Helmchen and colleagues of the object. Patients with cerebellar lesions have great diffi- used fMRI to find that activation in hemispheric lobule VI culty with such tasks [153]. Mice lacking genes associated with and in the anterior vermis varied with subject reports of pain cerebellar-dependent plasticity are selectively impairedinan intensity, though only when the stimuli were self-administered operant task that requires using precise sensory timing to restrict [146]. The authors suggested that these cerebellar regions movement latencies [154]. could reflect pain perception and are involved in signaling The preceding examples highlight a critical cerebellar role the expected sensory consequences of pain. In another study, in using sensory information to time movement. The reverse fMRI of trigeminal neuropathic pain elicited by brushing and situation, where movement is used to anticipate and modulate heat showed responses in crus I, crus II, and lobule VIIB that sensory information, is also cerebellar dependent, at least were not evoked by the non-painful control stimuli [147]. when the events are of a limited temporal extent. We have Recent neuroimaging evidence suggests that certain cere- general consensus that the cerebellum uses a forward model to bellar responses during pain may reflect multi-modal aversive generate a prediction of the expected sensory consequences of processing. An fMRI study found that noxious heat and the an action [155]. Kotz and colleagues [156]provide aparticu- passive viewing of unpleasant pictures activated overlapping larly compelling EEG example. The early N100 response regions of the cerebellum: hemispheric lobules VI, VIIB, and evoked by an auditory stimulus is markedly attenuated when crus I [148]. Further analysis revealed that these areas of the tones are triggered by a volitional action compared with functional overlap were significantly inversely correlated with when the tones are externally triggered. This attenuation is activation in the anterior hypothalamus, subgenual anterior essentially absent in patients with focal cerebellar lesions of cingulate cortex, and the parahippocampal gyrus. These find- the left or right hemisphere, with the sensory response similar ings suggest that responses in these cerebellar regions are not for self-triggered and externally triggered actions (see specific to pain processing but appear to apply to other aver- Fig. 6b). Forward models, as a form of prediction, have been sive sensory and affective experiences as well [149]. Howev- er, other functions related to pain aside from aversion may also employed to describe brain function more generally [157]. A be processed in the cerebellum, as areas that responded to challenge is to specify the conditions that distinguish noxious heat and not to aversive pictures were also identified cerebellar-dependent and cerebellar-independent forward including crus II. Further study is required to determine the models. One possibility is that, as with classical conditioning, functional topography of the cerebellum as it relates to pain the cerebellar domain is defined by temporal constraints, and its different sensory, affective, and cognitive components. situations in which the predictions require some form of precise temporal representation. In one oft-cited example, the tickling sensation from self-generated movements be- comes more intense when delays are introduced between the Sensory Processing and the Cerebellum: Timing action and the somatosensory stimulation [158]. Similarly, (R.B. Ivry) learning rates are dramatically reduced with delayed feedback during visuomotor adaptation [159]. Movement dynamically incorporates sensory information and The strongest evidence for a critical role of the cerebellum anticipates the sensory consequences of the action (see also in sensory timing comes from tasks that do not entail overt the section by Dr. Bower, “Is the Cerebellum Sensory for movement (see also the section by Drs. Baumann and Motor’s Sake, or Motor for Sensory’s Sake? (J.M. Bower)”). Mattingley, “The Role of the Cerebellum in Visual and Audi- While this is a general feature of motor control, there is tory Processing (O. Baumann and J.B. Mattingley)”). Re- consensus of a cerebellar dependency on tasks that impose search here falls into three general domains. First are tasks precise temporal constraints. A prominent feature of cerebellar examining how the cerebellum responds to temporal regular- ataxia is the loss of the fine temporal patterning that is char- ities, or perhaps more telling, violations of temporal expec- acteristic of skilled movement. Experimentally, eyeblink con- tancies. Tesche [160] compared evoked MEG responses to ditioning has proven to be an exquisite model system for periodic (predictable) tactile stimuli or epochs in which the Cerebellum (2015) 14:197–220 209 Fig. 6 Cerebellum and sensory timing. a Adaptive timing of conditioned eye blink response is abolished following infusion of picrotoxin, an agent that disrupts input from cerebellar cortex to deep cerebellar nuclei. Courtesy of Michael Mauk. b Patients with focal cerebellar lesions fail to show attenuated ERP response to self-generated sounds compared with externally produced sounds. Adapted from [156]. c Patients with cerebellar degeneration (SCA6) exhibit selective deficit on time perception tasks that require interval timing (Var, Fix) while spared performance on tasks that require beat-based timing (Reg, Iso, Met). Adapted from [165]. d Cerebellar grey matter volume is correlated with perceptual acuity on time discrimination task, relative to a color discrimination task. Adapted from [173] stimulus was withheld (prediction violations). Whereas the cerebellar BOLD response was larger in the latter compared evoked response in somatosensory cortex was stimulus- with the former. Converging evidence comes from a study locked and independent of predictability, the cerebellar re- showing that patients with cerebellar pathology are impaired sponse was anticipatory, leading the expected onset of the in adapting to velocity perturbations in this task [58]. stimulus. Moreover, it was markedly larger following a viola- Third, and perhaps most direct, are studies of duration tion, consistent with the idea that the cerebellum was sensitive discrimination. Ivry and Keele [88] provided the first evidence to temporal prediction violations. Further support for this idea of a “pure” sensory timing deficit in patients with cerebellar comes from fMRI work showing larger cerebellar activation pathology. The patients were impaired in judging the duration to visual stimuli with unpredictable timing (e.g., [161]) as well of an auditory stimulus but showed normal performance in as a study in which an early ERP signal to deviant auditory judging stimulus loudness. This finding has been confirmed in stimuli was found to be abnormal in patients with cerebellar various studies over the past 25 years, including one study in degeneration [162]. which testing was restricted to a large group of patients with The second domain involves studies of velocity perception. SCA6, a condition in which the pathology is relatively re- Cellular activity in the posterior cerebellum is sensitive to stricted to the cerebellar cortex [165](see Fig. 6c), and studies stimulus motion (see sections by Dr. Cullen on “The Cerebel- with healthy individuals in which cerebellar function has been lum and Perception: The Role of the Cerebellum in Self- transiently disrupted by TMS [166, 167]. There is general Motion Perception (K.E. Cullen)”, and Drs. Sokolov and consensus that cerebellar contributions to sensory (and motor) Pavlova, “Cerebellar Involvement in Biological Motion Pro- timing are most pronounced with relatively short intervals cessing (A.A. Sokolov and M.A. Pavlova)”). It is possible that (less than 1 s) and in the representation of intervals (either these signals are related to preparation of potential eye or body absolute or relative as in state estimation models) rather than movements. However, a causal contribution to perception more complex temporal relationships (e.g., rhythm). The few comes from psychophysical studies showing that patients with negative results on duration discrimination are also informa- cerebellar pathology are impaired in visual motion discrimi- tive: They have involved patients with unilateral lesions [168, nation [86, 163]. Moreover, the cerebellar contribution appears 169], suggesting that a single intact cerebellar hemisphere to be most critical when the motion perception task requires may be sufficient to support sensory timing [170]. The func- time-based judgments. O’Reilly [164] used a task in which a tional neuroimaging literature on duration perception has moving stimulus disappeared behind an occluder. When the proven more difficult to decipher [171], especially since many stimulus reappeared, the participant had to judge if there had studies do not provide adequate coverage of the cerebellum. Interestingly, three recent structural MRI studies report a been a deviation in direction (spatial) or speed (temporal). The 210 Cerebellum (2015) 14:197–220 positive correlation between measures of cerebellum volume Mismatch negativity (MMN) studies in subjects with cer- and temporal acuity in healthy individuals [172–174] ebellar damage in the somatosensory [182] or auditory [162] (see Fig. 6d). domain have confirmed this hypothesis. MMN is believed to This is not to say there is consensus for the uniqueness be generated by an automatic cortical change-detection pro- of the cerebellum in sensory timing. Indeed, there is con- cess that is activated by differences between current and prior sensus that the cerebellum is not the sole structure capable inputs. When the MMN protocol is applied to subjects with of representing temporal information. The challenge re- cerebellar lesions, the MMN response is absent or abnormal. mains to develop more analytic tasks and models that Per the long-standing model in which the cerebellum acts as a provide better specification of the various operations re- comparator [183], it has been proposed that, in the cerebellum, quired in tasks that require precise temporal processing. actual input and preceding stimuli are compared, and discor- Nonetheless, the cerebellar timing hypothesis [98]has dances are identified. If the incoming stimulus corresponds to proven to be of considerable utility for exploring the the predicted stimulus, cerebellar output is minimal; if a function, structure, and physiological of the cerebellum discrepancy–error signal is detected, the activity in the cere- in motor control and beyond. bellum increases and a large area of the cerebral cortex is alerted by enhancing its excitability (Fig. 7). We developed a “sequence detection model” to describe the operational mode of cerebellar processing not only in somato- The Cerebellum in Predicting Perceptual Events sensory [182], but also in visuospatial [184] and cognitive (M. Leggio and M. Molinari) domains [185]. Cerebellar patients were impaired specifically in the recognition of spatial sequences when tested on a visuo- Perception can be considered the result of interactions in time spatial serial reaction time task [184]. Results of visuospatial between a dynamic mind and a dynamic world. To achieve tests demonstrated that subjects with cerebellar damage were mind-world synchronization, our perceptual systems must impaired specifically with regard to sequence recognition, even constantly tune themselves to an ever-changing environment. to a greater extent than sequence execution [184]. Furthermore, Perceptual tuning, like the sensorimotor tuning that is needed by forcing the declarative knowledge of the spatial order, it was for smooth movement control, can be obtained only if predic- possible to improve performance significantly. Similar findings tion capabilities are embedded in the process [175]. Moreover, have been reported by several groups [186–191], supporting predictive processing represents a fundamental principle of cerebellar function in extracting sequential order information neural computations in the brain [176]. from incoming sensory information [184]. Many groups have attempted to identify the neural bases of Subjects with cerebellar damage also develop impairments in foresight, and despite considerable ongoing debate, a consen- cognitive sequencing [192]. We analyzed prediction ability in sus exists on the importance of the cerebellum in prediction patients with cerebellar damage who performed a cognitive task [177]. To make the matter even more interesting for cerebellar in which predictability was based primarily on abstract/spatial, scientists, data are accumulating on the significance of the behavioral/visual, or behavioral/linguistic sequence information cerebellum for sensory processing and in optimizing percep- [192]; in this task, sets of cartoon-like drawings that reproduced tion [58]. Perceptual optimization and prediction of incoming behavioral sequences were to be placed in the correct order. The sensory information have been suggested to be effected by patients were impaired in sequencing events in all domains, sequence processing in cerebellar circuits [58, 178, 179]. developing domain-related specificity, based on the side of the Using magnetoencephalographic recordings, Tesche and cerebellar lesion. Thus, although no specific sequencing locali- Karhu [160] demonstrated that cerebellar activity is enhanced zation can be identified, sequencing processing can be found in after an unpredictable omission is inserted into a regular train the different cerebellar functional domains. This impairment of somatosensory stimuli. As a result, no activity is present in suggests difficulties in perceiving the depicted behavior correctly. the parietal cortex, whereas a notable response develops in the This evidence is consistent with difficulties that are encountered cerebellum. Consequently, it can be argued that the cerebel- in tuning behavior and the environment correctly not only after lum detects the absence of a somatosensory stimulus to a cerebellar damage [19], but also in behavioral pathologies, such greater extent than its presence. This response to the absence as autism and schizophrenia, disorders that have been linked to of a stimulus can be understood only as an indication that cerebellar abnormality [193, 194]. something that is expected does not appear [180]. If a sensory The hypothesis that pattern detection and prediction repre- pattern is recognized, it is possible to predict the sequence of sent a specific role in cerebellar function in perception is events and consequently anticipate each one [181]. Thus, in appealing, and compelling data from various sources support predicting incoming sensory information, the cerebellum gov- the sequence detection model of impaired cerebellar percep- erns the detection of the absence of an expected stimulus and tion. Furthermore, the perceptual deficits that are observed in the appearance of an unexpected stimulus. schizophrenia [195, 196] and autism [197, 198] resemble Cerebellum (2015) 14:197–220 211 Fig. 7 Sequence detection model of prediction. If sensory events appear in a fixed sequence repeatedly in a short time, the sensory sequence is implicitly memorized a which allows cerebellar circuits to compute a prediction for forthcoming perceptual events b.If the prediction holds c, a signal is sent to the cerebral cortex to alert selective brain areas, which become activated prior to the realized event and are thus better suited to process the incoming stimulus. If the prediction fails d, an alertsignalis sent, andbrain activation is more widespread, accelerating the processing of salient sensory information by the changing events and attuning the behavioral response to the new environment cerebellar dysfunctions. Notably, cerebellar pathogenic mech- behavioral including human studies. First in a series of imag- anisms have been hypothesized to mediate schizophrenia ing experiments, we demonstrated, as the hypothesis predicts, [199] and autism [200], and the existence of cerebellar-like that activity in the human cerebellum [51, 206] and related sequence detection deficits [201, 202] is additional support for structures [207] is substantially greater when fingers are used the cerebellar pathogenic theories of these diseases. in a tactile discrimination task. A meta-analysis of neuroim- aging data then generalized this result to the auditory system, suggesting larger and more spatially extensive activations Is the Cerebellum Sensory for Motor’s Sake, or Motor during discriminative auditory tasks [208], a result subse- for Sensory’s Sake? (J.M. Bower) quently confirmed using PET [90]. Importantly, the PETstudy also supported a further important prediction of the sensory acquisition hypothesis, namely that cerebellar activity should This title is the same as a paper published more than 15 years ago describing our hypothesis that the cerebellum controls the increase with task difficulty, i.e., when better control of the quality of sensory data is likely more important [97]. A similar acquisition of sensory data [203], an idea first proposed even 10 years earlier in a paper exploring the spatial structure of the result has been reported independently in a combined human visual and auditory imaging study [89]. extensive peri-oral tactile representations in the cerebellum of rats: While human imaging data can be suggestive of brain “… we suggest that tactile regions of the cerebellum are function, an important test of any functional hypothesis is its involved in controlling the movements specifically associated ability to predict behavioral results. In this case, it has long with active tactile exploration …to coordinate the use of been a central tenet of cerebellar descriptions that the structure sensory structures so that the highest quality sensory informa- has no influence on sensory perception [209]. However, be- tion is being obtained by the rest of the nervous system during cause sensory perception is based on the quality of sensory data, we predicted that impairment of cerebellar function the exploratory process. By monitoring the acquisition of sensory information and adjusting motor performance accord- should have sensory perceptive consequences [203]. Consis- tent with this prediction, we have shown that humans with ingly, cerebellar circuits would be expected to substantially improve the efficiency of sensory processing by the rest of the cerebellar degenerative disease have significantly poorer thresholds for pitch discrimination [57]. Other studies in au- nervous system.” (p. 776, [204]) dition [86], somatosensation [210, 211], proprioception, and vision [212] have now also demonstrated cerebellar-related Evidence in Support primary sensory deficits, which have also been reported using higher order tasks like speech [213], motion detection [214], While a model-based re-analysis of cerebellar cortical net- works alsosupportsthishypothesis[205], this review will analysis of temporal sequences [178], as well as the general perception of time [167, 215]. focus on supporting experimental results from more 212 Cerebellum (2015) 14:197–220 Finally, while human psychophysical and imaging studies, but instead because sensory information is temporally coded properly designed, can test functional hypotheses, linking at the neuronal level [228], and therefore experimental these hypotheses to actual physical computational mecha- manipulations of expected timing relationships in presented nisms still requires the use of animal models [205]. While stimuli are likely to evoke stronger cerebellar effects. the large majority of animal studies exploring the functional 3. The cerebellum is invoked in proportion to the need for significance of the extensive sensory projections to the cere- sensory vigilance. Another important prediction of our bellum continue to frame the results in the context of tradi- hypothesis is that cerebellar involvement will scale as better tional motor control theories [216], a recent behavioral study controlled sensory data are required [90], making it impor- in rats has demonstrated that optogenetic stimulation of the tant to evaluate task difficulty when considering cerebellar- cerebellum specifically disrupts the use of the whiskers during related sensory effects [58, 229, 230]. Interestingly, numer- active touch [217]. These authors specifically conclude that ous cerebellar studies already employ masking sensory noise their results support a role of the cerebellum in the “optimiza- to evoke larger cerebellar responses [89, 92]orreveal be- tion of sensory data acquisition” (p. 6, [217]). havioral deficits [163]. Overcoming the consequences of sensory noise either applied externally or self-generated Implication for Theories of Cerebellar Sensory Function [231] we predict will especially increase requirements for cerebellar control. This effect also confounds the interpreta- With growing evidence that the cerebellum plays some role in tion of sensory stimuli like pain [141], whichontheir own sensory function, it is time to fully reconsider cerebellar increase subject vigilance [148, 232], as well as studies of function from a sensory point of view: mechanisms like attention [233–235]. 4. The cerebellum is a support structure. Perhaps the most 1. Re-interpreting cerebellar involvement in motor control. important implication of the sensory hypothesis is that the It has been known for more than 150 years that lesions of cerebellum performs a more internal than external func- the cerebellum disrupt movement [218, 219] with the tion. Instead of itself contributing directly to sensory majority of cerebellar theories accordingly focused on perception, the influence of the cerebellum is predicted mechanisms of direct motor control [220]. In contrast, to be indirect, facilitating the computational efficiency of the sensory data acquisition hypothesis proposes that the rest of the brain, including cerebral cortex [163]. To cerebellar effects on movement are an indirect conse- quote again from 25 years ago: quence of disrupting the sensory data on which motor behavior depends [97]. This prediction is consistent with “It has been largely accepted that the flocculus of the recent evidence that cerebellar patients have difficulty cerebellum is involved in adjusting the gain of the discriminating proprioceptive stimuli [210] and that a vestibulo-occular reflex to assure a minimal slip of significant component of cerebellar ataxia results from images on the retina during head movement [236, the inability of patients to perceive environmental insta- 237]. Psychophysical experiments demonstrate that bilities [221]. For this reason, it is critically important that more than 3°/sec of retinal slip starts to significantly motor-related studies, perhaps especially those involving degrade visual acuity and thus the ability of the visual purported motor learning [222], control for cerebellar system to process sensory information [238]. Thus the effects on primary sensory data. proposed role of the cerebellum, in VOR control, is to 2. Removing the legacy of cerebellar motor control theories. assure that the highest possible quality of visual infor- At present, most explanations for cerebellar involvement mation is provided to the visual system. In principle, this in non-motor-related behaviors assume that evolution has role is analogous to the role we are suggesting for lateral adopted cerebellar motor control computational mecha- tactile regions of the cerebellar cortex.” (p. 776, [204]). nisms to non-motor tasks [58, 73, 92, 211, 220, 223–226], including, for example, a presumed general role for the For sensory systems like vision, audition, olfaction, and cerebellum in timing not only of muscle activations dur- somatosensation, which in humans involve the largest part of ing movement but also of sensory perception [98, 227]. the cerebellum, the ‘support system’ status of cerebellum also While our analysis of cerebellar cortical circuitry ques- suggests a different interpretation of the important relationship between the cerebellum and the cerebral cortex. While in the tions the circuitry-based evidence for the original timing hypothesis [205], we do expect that any disruption in traditional motor-control context, the influence of the cerebral cortex on the cerebellum is generally described as sensory data acquisition control may very well be partic- ularly apparent with tasks involving precise timing (see implementing a kind of forward model to (quoting Dr Ivry in this article) “generate a prediction of the expected sensory the section by Dr. Ivry, “Sensory Processing and the Cerebellum: Timing (R.B. Ivry)”). This is not, however, consequences of an action” (see also the section by Dr. Paulin, because the cerebellum itself implements a timing function, “Evolutionary Perspectives on Cerebellar Function (M.G. Cerebellum (2015) 14:197–220 213 Paulin)”), in our view, the influence of the cerebral cortex on Summary and Conclusions the cerebellum provides contextual information related to the expected use of the sensory data by cerebral cortex. We don’t The aim of this consensus paper is to capture the range of think that such a function explicitly involves a ‘prediction’ as experimental approaches and theoretical models that have much as it does a continuous stream of contextual informa- contributed to our current understanding of the influence of tion. In fact, although again beyond the scope of the current the cerebellum on perceptual processes. Contributions from commentary, our analysis of cerebellar cortex suggests that its fourteen experts, spanning a range of methodological ap- circuitry specifically places information arising from particu- proaches and with different theoretical views, have been lar sensory receptors (e.g., the upper lip) in the context of other brought together to provide an up-to-date snapshot of thinking sensory surfaces involved at the same time in sensory data on this topic. acquisition (e.g., the lower lip). We have proposed that cere- The outcome of this project indicates that no single, coher- bellar output (through direct projections to the midbrain and ent model has yet emerged regarding the mechanisms by brain stem motor centers as well as potentially through motor which the cerebellum may influence perception. Nonetheless, regions of cerebral cortex) then makes subtle relative adjust- it is important to assemble the empirical data, showing the ments in the position of tactile sensory surfaces to optimize the association of the cerebellum with a wide range of perceptual information content. A recent analysis of the influence of the systems including those related to vision, audition, touch, cerebellum on whisking in rats supports this prediction [217]. proprioception, self-motion perception, and nociception. The Similarly, we have proposed that the cerebellum also likely possible anatomical and physiological underpinnings of this modulates the cochlear outer hair cells during auditory data broad influence was reviewed by Dr. Schmahmann, acquisition. In fact, we have suggested that the cerebellum documenting significant cerebellar connection with sensory, plays the same role for all sensory systems. as well as associative and paralimbic, areas of the cerebrum. 5. Implications for human disease. Finally, the most exciting These findings are corroborated by human neuroimaging application of this sensory focused hypothesis may be to studies, which show that fMRI resting-state signals in the human health and disease. Although understudied, it has cerebellum correlate significantly with those in visual and been known for more than 150 years that motor control auditory cortices in the cerebrum (see the section by Dr. can recover after cerebellar cortical lesions [239, 240]an Habas, “Resting-State Functional Connectivity Between Cer- effect also now demonstrated for presumed ‘cognitive’ ebellum and Sensory Systems (C. Habas)”). Second, a number function [241–243]. The sensory hypothesis attributed of the commentators described clinical studies that show how this recovery to the eventual adaptation of the rest of the cerebellar lesions can lead to deficits in a diverse set of brain to less well-controlled sensory data [203]. Evidence perceptual tasks, including visual motion perception, auditory has also been growing that the cerebellum plays a role in pitch perception, self-motion perception, biological motion autism spectrum disorders (ASD), although there is no perception of others, time perception, and the recognition of consensus for the mechanism [244]. In the context of our perceptual sequences (see sections by Drs. Baumann and hypothesis, the relationship is quite direct, with ASD seen Mattingley, “The Role of the Cerebellum in Visual and Audi- as a behavioral adaptation to a general and overwhelming tory Processing (O. Baumann and J.B. Mattingley)”; Drs. lack of control over the process of sensory data acquisi- Pavlova and Sokolov, “Cerebellar Involvement in Biological tion. From this perspective, therapies that focus on repet- Motion Processing (A.A. Sokolov and M.A. Pavlova)”;Dr. itive behaviors in highly controlled sensory environments Cullen, “The Cerebellum and Perception: The Role of the with specific emphasis on sensory integration [245] Cerebellum in Self-Motion Perception (K.E. Cullen)”;Dr. would, we suggest, establish sensory conditions making Ivry, “Sensory Processing and the Cerebellum: Timing (R.B. it easier for the brain to learn to compensate for the lack of Ivry)”; Drs. Leggio and Molinari, “The Cerebellum in stable sensory data. It may even be worth considering Predicting Perceptual Events (M. Leggio and M. Molinari)”; whether the apparent increasing incidence of ASD could and Dr. Bower, “Is the Cerebellum Sensory for Motor’s Sake, be attributable to sensory over-stimulation of children or Motor for Sensory’s Sake? (J.M. Bower)”). Third, human before the late developing cerebellum is fully functional. neuroimaging studies have consistently shown reliable cere- In summary, there is no question that the evidence is bellar activation during performance of a range of perceptual growing for some kind of cerebellar involvement in tasks, independent of any motor-related activity of observers mechanisms of sensory function. However, instead of (see sections by Drs. Baumann and Mattingley, “The Role of assuming a direct role in these mechanisms borrowing the Cerebellum in Visual and Auditory Processing (O. traditional cerebellar theories designed to explain motor Baumann and J.B. Mattingley)”; Drs. Pavlova and Sokolov, control, in our view, this new evidence should instead call “Cerebellar Involvement in Biological Motion Processing into question the historical view of the cerebellum as (A.A. Sokolov and M.A. Pavlova)”; Drs. Borra and Moulton, primarily a motor control device. “Pain and the Cerebellum (R.J. Borra and E.A. Moulton)”; 214 Cerebellum (2015) 14:197–220 and Dr. Bower, “Is the Cerebellum Sensory for Motor’s Sake, these hypotheses. Most of the current evidence is deliv- or Motor for Sensory’s Sake? (J.M. Bower)”). ered by human lesion and neuroimaging studies, methods In summary, it seems the answer to the question of whether that have provided valuable insights from a systems-level the cerebellum plays a role in perception is unequivocally perspective, but are of limited value in constraining affirmative. What remains to be determined is precisely how models at the level of microcircuitry. It is therefore essen- the cerebellum contributes to perceptual processes. tial to also explore the cerebellum’s involvement in per- Dr. Schmahmann sets the stage for functional hypotheses. ceptual tasks at the level of single neurons. Dr. Cullen’s Inspired by the cerebellum’s uniform neuroanatomical struc- research on the role of the cerebellum in self-motion ture and dense heterogeneous connectivity, he argues that we perception provides a compelling example. By recording should assume a constant computation—the universal cere- from individual cerebellar neurons, her research has shown bellar transform—that is applied to multiple domains of neu- that the cerebellum computes sensory prediction error sig- rological function determined by cerebellar connections. The nals that effectively distinguish between the sensory con- idea of a uniform computation is repeated in many of the other sequences of self-generated and externally produced ac- commentaries, although the specific form of the computation tions. These findings seem inconsistent with the conven- shows considerable variation. Building on comparative data tional view that the role of the cerebellum is restricted to from across the animal kingdom, Dr. Paulin suggests that the motor learning. cerebellum provides the ability to predict state trajectories of Finally, an important application of new knowledge arising dynamical systems. The ability to predict state trajectories of from research into the role of the cerebellum in perception is in the body and external targets is essential for agile motor the domain of human health and disease. The historical asso- control and can explain the obvious, classical symptoms of ciation of the cerebellum with “motor function” has limited cerebellar dysfunction. But state estimation can also provide appropriate consideration of its potential role in perceptual core capability for a variety of signal processing, decision- functions, in both health and disease. It is now apparent that making and control tasks, and this could explain newer evi- cerebellar lesions can lead to a range of behavioral, cognitive, dence about the cerebellum’s role in non-motor tasks. The affective, and perceptual impairments. In addition, psychiatric latest neuroimaging evidence for direct interaction between conditions that are characterized by perceptual and cognitive the cerebellum and temporal areas involved in visual motion (as well as motor) disturbances, including autism, schizophre- processing and body motion processing (MT/MST and STS), nia, and attention deficit hyperactivity disorder, are associated as presented by Drs. Baumann, Mattingley, Pavlova and with cerebellar pathology. The possibility of a cerebellar role Sokolov, appears to lend further support to this hypothesis. in the manifestations or pathogenesis of these conditions Similarly, Dr. Ivry’s hypothesis proposes a contribution of the is intriguing. Further research into the role of the cere- bellum in perceptual functions may help to advance our cerebellum to the analysis and prediction of sensory event timing in the sub-second range. Drs. Leggio and Molinari’s understanding of the mechanisms underlying these dis- hypothesis of the cerebellum’s role in perception shares the orders. Moreover, patients with isolated cerebellar in- central assumption that the cerebellum is involved in the sults, cerebellar tumors, and hereditary cerebellar degen- analysis and prediction of dynamic perceptual events. While erative disease will also benefit from a better under- Dr. Ivry focused here on a narrower view of prediction, events standing of the role of the cerebellum in perception. requiring precise timing in the sub-second range, Drs. Leggio To date, diagnostic evaluation and therapeutic interven- and Molinari take a broader view of prediction with their tions in patients with cerebellar disease have been lim- hypothesis that the cerebellum supports perception by ited to the striking deficits in the coordination of vol- extracting sequential order information from incoming senso- untary movements. Recognition of a cerebellar role in ry information. Clinical and neuroimaging studies not sensory processes helps to identify and treat potential only implicate the cerebellum in the analysis of dynam- perceptual deficits that may at present go unnoticed and ic stimuli, but also in less dynamic perceptual tasks untreated. In addition, further research on the compen- such as pitch discrimination and nociception. Dr. Bower satory potential of not only motor, but also perceptual urges us to consider that the cerebellar contribution cerebro-cerebellar networks after cerebellar damage may arises at an even earlier stage of processing, arguing advance both clinical management and understanding of that the cerebellum influences perception by controlling the cerebellar contribution to perception. the acquisition of sensory data, an idea that might This review is the first attempt to capture the variety of explain why cerebellar activity often increases with the current experimental approaches and theoretical models on difficulty of a perceptual task. the cerebellum’s role or influence on perception. By drawing While some of the described theories could be seen as together the diverse perspectives, we intend to stimulate sci- complementary, the challenge remains to develop more entific debate and increase interest in the cerebellum and its explicit experimental tests that can distinguish between complex functions. Cerebellum (2015) 14:197–220 215 Acknowledgments (1) Dr. Schmahmann’s work was supported in part 14. Schmahmann JD, Pandya DN. Projections to the basis pontis from by the MINDlink and Birmingham Foundations. (2) Dr. Baumann was the superior temporal sulcus and superior temporal region in the supported by an Australian Research Council (ARC) Discovery Early rhesus monkey. J Comp Neurol. 1991;308:224–48. Career Researcher Award (DE120100535) and Dr. Mattingley by an 15. Ungerleider LG, Desimone R, Galkin TW, Mishkin M. Subcortical ARC Australian Laureate Fellowship (FL110100103), the ARC-SRI projections of area MT in the macaque. J Comp Neurol. 1984;223: Science of Learning Research Centre (SR120300015), and the ARC 368–86. Centre of Excellence for Integrative Brain Function (ARC Centre Grant 16. Schmahmann JD, Pandya DN. Prelunate, occipitotemporal, and CE140100007). (3) Dr. Pavlova was supported by Else Kröner Fresenius parahippocampal projections to the basis pontis in rhesus monkey. Foundation (Grant P2013_127), the Reinhold-Beitlich Foundation, the J Comp Neurol. 1993;337:94–112. Berthold Leibinger Foundation, and the Heidehof Foundation (Grant 17. Fiez JA, Raichle ME. Linguistic processing. In: Schmahmann JD, 59073.01.1/3.13). (4) Dr. Borra was supported by funding from the Sigrid editor. The cerebellum and cognition. International review of neu- Juselius Foundation, the Instrumentarium Research Foundation, the Finn- robiology, vol. 41. San Diego: Academic; 1997. p. 233–54. ish Medical Foundation, the Paulo Foundation and the Academy of 18. Stoodley CJ, Schmahmann JD. Functional topography in the human Finland (270352). Dr. Moulton was supported by the National Institute cerebellum: a meta-analysis of neuroimaging studies. Neuroimage. of Health, USA (NIH/NCI R21CA185870). (5) Dr. Ivry was supported 2009;44:489–501. by the National Institute of Health, USA (NS084948 and NS074917). 19. Schmahmann JD, Sherman JC. The cerebellar cognitive affective syndrome. Brain. 1998;121:561–79. 20. Mariën P, Ackermann H, Adamaszek M, et al. Consensus paper: Conflicts of Interest The authors declare that no financial or personal language and the cerebellum: an ongoing enigma. Cerebellum. competing interests exist. 2014;13:386–410. 21. Picard N, Strick PL. Motor areas of the medial wall: a review of their Open Access This article is distributed under the terms of the Creative location and functional activation. Cereb Cortex. 1996;6:342–53. Commons Attribution License which permits any use, distribution, and 22. Vilensky JA, Van Hoesen GW. Corticopontine projections from the reproduction in any medium, provided the original author(s) and the cingulate cortex in the rhesus monkey. Brain Res. 1981;205:391–5. source are credited. 23. Aas J-E, Brodal P. Demonstration of topographically organized projections from the hypothalamus to the pontine nuclei: an exper- imental study in the cat. J Comp Neurol. 1988;268:313–28. 24. Haines DE, Dietrichs E. An HRP study of hypothalamo-cerebellar References and cerebello-hypothalamic connections in squirrel monkey (Saimiri sciureus). J Comp Neurol. 1984;229:559–75. 25. Ramnani N, Behrens TE, Johansen-Berg H, et al. The evolution of 1. Schmahmann JD. From movement to thought: anatomic substrates prefrontal inputs to the cortico-pontine system: diffusion imaging of the cerebellar contribution to cognitive processing. Hum Brain evidence from Macaque monkeys and humans. Cereb Cortex. Mapp. 1996;4:174–98. 2006;16:811–8. 2. Schmahmann JD. The cerebrocerebellar system: anatomic sub- 26. Allen GI, Tsukahara N. Cerebrocerebellar communication systems. strates of the cerebellar contribution to cognition and emotion. Int Physiol Rev. 1974;54:957–1008. Rev Psychiatry. 2001;13:247–60. 27. Brodal P. The pontocerebellar projection in the rhesus monkey: an 3. Schmahmann JD. The role of the cerebellum in cognition and experimental study with retrograde axonal transport of horseradish emotion: personal reflections since 1982 on the dysmetria of peroxidase. Neuroscience. 1979;4:193–208. thought hypothesis, and its historical evolution from theory to 28. Kelly RM, Strick PL. Cerebellar loops with motor cortex and therapy. Neuropsychol Rev. 2010;20:236–60. prefrontal cortex of a nonhuman primate. J Neurosci. 2003;23: 4. Schmahmann JD, Pandya DN. The cerebrocerebellar system. In: 8432–44. Schmahmann JD, editor. The cerebellum and cognition. San Diego: 29. Middleton FA, Strick PL. Cerebellar output channels. In: Schmahmann Academic; 2007. p. 31–60. JD, editor. The cerebellum and cognition. International review of 5. Stein JR, Glickstein M. Role of the cerebellum in visual guidance of neurobiology, vol. 41. San Diego: Academic; 1997. p. 61–82. movement. Physiol Rev. 1992;72:967–1017. 30. Buckner RL, Krienen FM, Castellanos A, Diaz JC, Yeo BT. The 6. Oscarsson O. Functional organization of the spino- and organization of the human cerebellum estimated by intrinsic func- cuneocerebellar tracts. Physiol Rev. 1965;45:495–522. tional connectivity. J Neurophysiol. 2011;106:2322–45. 7. Barmack NH. Central vestibular system: vestibular nuclei and pos- 31. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connec- terior cerebellum. Brain Res Bull. 2003;60:511–41. tivity in the motor cortex of resting human brain using echo-planar 8. Brodal P. The corticopontine projection in the rhesus monkey. MRI. Magn Res Med. 1995;34:537–41. Origin and principles of organization. Brain. 1978;101:251–83. 32. Beckmann CF, DeLuca M, Devlin JT, Smith SM. Investigations 9. Glickstein M, May JG, Mercier BE. Corticopontine projection in into resting-state connectivity using independent component analy- the macaque: the distribution of labeled cortical cells after large sis. Philos Trans R Soc Lond B Biol Sci. 2005;360:1001–13. injections of horseradish peroxidase in the pontine nuclei. J Comp 33. Krienen FM, Buckner RL. Segregated fronto-cerebellar circuits re- Neurol. 1985;235:343–59. vealed by intrinsic functional connectivity. Cereb Cortex. 2009;19: 10. Schmahmann JD, Pandya DN. Anatomical investigation of projec- 2485–97. tions to the basis pontis from posterior parietal association cortices 34. O’Reilly JX, Beckmann CF, Tomassini V, Ramnani N, Johansen-Berg in rhesus monkey. J Comp Neurol. 1989;289:53–73. H. Distinct and overlapping functional zones in the cerebellum defined 11. Schmahmann JD, Rosene DL, Pandya DN. Motor projections to the by resting state functional connectivity. Cereb Cortex. 2009;20:953–65. basis pontis in rhesus monkey. J Comp Neurol. 2004;478:248–68. 35. Habas C, Kamdar N, Nguyen D, Prater K, Beckmann CF, Menon V, 12. Schmahmann JD, MacMore J, Ko R. The human basis pontis. et al. Distinct cerebellar contributions to intrinsic connectivity net- Clinical syndromes and topographic organization. Brain. works. J Neurosci. 2009;29:8586–94. 2004;127:1269–91. 36. Stoodley CJ, Schmahmann JD. Evidence for topographic organiza- 13. Schmahmann JD, Pandya DN. Anatomic organization of the basilar tion in the cerebellum of motor control versus cognitive and affec- pontine projections from prefrontal cortices in rhesus monkey. J tive processing. Cortex. 2010;46:831–44. Neurosci. 1997;17:438–58. 216 Cerebellum (2015) 14:197–220 37. Sokolov AA, Erb M, Grodd W, Pavlova MA. Structural loop 62. Young JZ. Cerebellum and control of eye-movements in cephalo- between the cerebellum and the superior temporal sulcus: evidence pods. Nature. 1976;264:572–4. from diffusion tensor imaging. Cereb Cortex. 2014;24:626–32. 63. Bell CC, Han V, Sawtell NB. Cerebellum-like structures and their 38. Habas C, Guillevin R, Abanou A. Functional connectivity of the implications for cerebellar function. Annu Rev Neurosci. 2008;31: superior human temporal sulcus in the brain resting state at 3 T. 1–24. Neuroradiology. 2011;53:129–40. 64. Montgomery JC, Bodznick D, Yopak KE. The cerebellum and 39. Sang L, Qin W, Liu Y, Han W, Zhang Y, Jiang T, et al. Resting-state cerebellum-like structures of cartilaginous fishes. Brain Behav functional connectivity of the vermal and hemispheric subregions of Evol. 2012;80:152–65. cerebellum with both cerebral and cortical networks and subcortical 65. Oertel D, Young ED. What’s a cerebellar circuit doing in the structures. Neuroimage. 2012;61:1213–25. auditory system? Trends Neurosci. 2004;27:104–10. 40. Ding K, Liu Y, Yan X, Lin X, Jiang T. Altered functional connec- 66. Farris SM. Are mushroom bodies cerebellum-like structures? tivity of the primary visual cortex in subjects with amblyopia. Arthropod Struct Dev. 2011;40:368–79. Neural Plast. 2013;2013:612086. 67. Zeil J. Visual homing: an insect perspective. Curr Opin Neurobiol. 41. Seeley WW, Menon V, Schatzberg AF, Keller J, Glover GH, Kenna 2012;22:285–93. H, et al. Dissociable intrinsic connectivity networks for salience 68. Purcell EM. Life at low Reynolds number. Am J Phys. 1977;45:3– processing and executive control. J Neurosci. 2007;27:2349–56. 11. 42. Baumann O, Mattingley JB. Functional topography of primary 69. Gould SJ. The structure of evolutionary theory. Cambridge: emotion processing in the human cerebellum. Neuroimage. Belknap Press; 2002. 2012;61:805–11. 70. Brown RG, Hwang PYC. Introduction to random signals and ap- 43. Habas C, Cabanis EA. Dissociation of the neural networks recruited plied Kalman filtering. 3rd ed. New York: Wiley; 1996. during a haptic object-recognition task: complementary results with 71. Haykin S. Kalman filtering and neural networks. New York: Wiley; a tensorial independent component analysis. AJNR Am J 2001. Neuroradiol. 2008;29:1715–21. 72. Doucet A, De Freitas N, Gordon N. Sequential Monte Carlo 44. Holmes G. Clinical symptoms of cerebellar disease—and their methods in practice. New York: Springer; 2001. interpretation. Lancet. 1922;2:59–65. 73. Paulin MG. Evolution of the cerebellum as a neuronal ma- 45. Schmahmann J. The cerebellum and cognition. International review chine for Bayesian state estimation. J Neural Eng. 2005;2: of neurobiology, vol. 41. San Diego: Academic; 1997. 219–34. 46. Bullock TH. Electroreception. Ann Rev Neurosci. 1982;5:121–70. 74. Paulin MG, Hoffman LF. Bayesian head state prediction: computing 47. Llinas RR. Cortex of cerebellum. Sci Am. 1975;232:56–71. the dynamic prior with spiking neurons. Shanghai: International 48. Nieuwenhuys R, Pouwels E. A light and EM study of lobe C1 of the Conference on Natural Computation ICNC; 2011. gigantocerebellum of the mormyrid fish Gnathonemus petersi.Anat 75. Buesing L, Bill J, Nessler B, Maass W. Neural dynamics as sam- Rec. 1973;175:399. pling: a model for stochastic computation in recurrent networks of 49. Paulin MG. The role of the cerebellum in motor control and per- spiking neurons. PLoS Comput Biol. 2011;7:11. ception. Brain Behav Evol. 1993;41:39–50. 76. Deneve S. Making decisions with unknown sensory reliability. 50. Bower JM, Parsons LM. Rethinking the lesser brain. Sci Am. Front Neurosci. 2012;6:75. 2003;289:50–7. 77. Franklin DW, Wolpert DM. Computational mechanisms of senso- 51. Gao JH, Parsons LM, Bower JM, Xiong JH, Li JQ, Fox PT. rimotor control. Neuron. 2011;72:425–42. Cerebellum implicated in sensory acquisition and discrimination 78. Knill DC, Pouget A. The Bayesian brain: the role of uncertainty in rather than motor control. Science. 1996;272:545–7. neural coding and computation. Trends Neurosci. 2004;27:712–9. 52. Hore J, Watts S. Skilled throwers use physics to time ball release to 79. Kording KP, Wolpert DM. Bayesian integration in sensorimotor the nearest millisecond. J Neurophysiol. 2011;106:2024–33. learning. Nature. 2004;427:244–7. 53. Klier EM, Angelaki DE. Gaze Stabilization and the VOR. In: Koob 80. Laurens J, Droulez J. Bayesian processing of vestibular information. GF, Le Moal M, Thompson RF, editors. Encyclopedia of behavioral Biol Cybern. 2007;96:389–404. neuroscience. Oxford: Academic Press; 2010. p. 569–75. 81. Lochmann T, Deneve S. Neural processing as causal inference. Curr 54. Kurtzer I, Trautman P, Rasquinha RJ, Bhanpuri NH, Scott SH, Opin Neurobiol. 2011;21:774–81. Bastian AJ. Cerebellar damage diminishes long-latency responses 82. Ma WJ, Beck JM, Latham PE, Pouget A. Bayesian inference with to multijoint perturbations. J Neurophysiol. 2013;109:2228–41. probabilistic population codes. Nat Neurosci. 2006;9:1432–8. 55. Miall RC, King D. State estimation in the cerebellum. Cerebellum. 83. Wolpert DM. Probabilistic models in human sensorimotor control. 2008;7:572–6. Hum Mov Sci. 2007;26:511–24. 56. Molinari M, Restuccia D, Leggio MG. State estimation, response 84. Schmahmann JD. An emerging concept. The cerebellar contribution prediction, and cerebellar sensory processing for behavioral control. to higher function. Arch Neurol. 1991;48:1178–87. Cerebellum. 2009;8:399–402. 85. Snider RS, Stowell A. Receiving areas of the tactile, auditory, and 57. Parsons LM, Petacchi A, Schmahmann JD, Bower JM. Pitch dis- visual systems in the cerebellum. J Neurophysiol. 1944;7:331–57. crimination in cerebellar patients: evidence for a sensory deficit. 86. Ivry RB, Diener HC. Impaired velocity perception in patients with Brain Res. 2009;1303:84–96. lesions of the cerebellum. J Cogn Neurosci. 1991;3:355–66. 58. Roth MJ, Synofzik M, Lindner A. The cerebellum optimizes per- 87. Thier P, Haarmeier T, Treue S, Barash S. Absence of a common ceptual predictions about external sensory events. Curr Biol. functional denominator of visual disturbances in cerebellar disease. 2013;23:930–5. Brain. 1999;122:2133–46. 59. Zago M, McIntyre J, Senot P, Lacquaniti F. Visuo-motor coordina- 88. Ivry RB, Keele SW. Timing functions of the cerebellum. J Cogn tion and internal models for object interception. Exp Brain Res. Neurosci. 1989;1:136–52. 2009;192:571–604. 89. Baumann O, Mattingley JB. Scaling of neural responses to visual 60. Hobbs MJ, Young JZ. Cephalopod cerebellum. Brain Res. 1973;55: and auditory motion in the human cerebellum. J Neurosci. 2010;30: 424–30. 4489–95. 61. Williamson R, Chrachri A. A model biological neural network: the 90. Petacchi A, Kaernbach C, Ratnam R, Bower JM. Increased activa- cephalopod vestibular system. Philos Trans R Soc Lond B Biol Sci. tion of the human cerebellum during pitch discrimination: a positron 2007;362:473–81. emission tomography (PET) study. Hear Res. 2011;282:35–48. Cerebellum (2015) 14:197–220 217 91. Sokolov AA, Gharabaghi A, Tatagiba MS, Pavlova M. Cerebellar 113. Blake R, Turner LM, Smoski MJ, Pozdol SL, Stone WL. Visual engagement in an action observation network. Cereb Cortex. recognition of biological motion is impairedinchildrenwith autism. 2010;20:486–91. Psychol Sci. 2003;14:151–7. 92. Sokolov AA, Erb M, Gharabaghi A, Grodd W, Tatagiba MS, 114. Kim J, Doop ML, Blake R, Park S. Impaired visual recognition of Pavlova MA. Biological motion processing: the left cerebellum biological motion in schizophrenia. Schizophr Res. 2005;77:299– communicates with the right superior temporal sulcus. 307. Neuroimage. 2012;59:2824–30. 115. Barnea-Goraly N, Kwon H, Menon V, Eliez S, Lotspeich L, Reiss 93. Baumann O, Greenlee MW. Neural correlates of coherent audiovi- AL. White matter structure in autism: preliminary evidence from sual motion perception. Cereb Cortex. 2007;17:1433–43. diffusion tensor imaging. Biol Psychiatry. 2004;55:323–6. 94. Bushara KO, Grafman J, Hallett M. Neural correlates of auditory– 116. Kanaan RA, Borgwardt S, McGuire PK, Craig MC, Murphy DG, visual stimulus onset asynchrony detection. J Neurosci. 2001;21: Picchioni M, et al. Microstructural organization of cerebellar tracts 300–4. in schizophrenia. Biol Psychiatry. 2009;66:1067–9. 95. Calvert GA, Hansen PC, Iversen SD, Brammer MJ. Detection of 117. Kaiser MD, Hudac CM, Shultz S, Lee SM, Cheung C, Berken AM, audio-visual integration sites in humans by application of electro- et al. Neural signatures of autism. Proc Natl Acad Sci U S A. physiological criteria to the BOLD effect. Neuroimage. 2001;14: 2010;107:21223–8. 427–38. 118. Catani M, Jones DK, Daly E, Embiricos N, Deeley Q, Pugliese L, 96. Cerminara NL, Apps R, Marple-Horvat DE. An internal model of a et al. Altered cerebellar feedback projections in Asperger syndrome. moving visual target in the lateral cerebellum. J Physiol. 2009;587: Neuroimage. 2008;41:1184–91. 429–42. 119. Nawrot M, Rizzo M. Motion perception deficits from midline 97. Bower JM. Control of sensory data acquisition. In: Schmahmann cerebellar lesions in human. Vision Res. 1995;35:723–31. JD, editor. The cerebellum and cognition. International review of 120. Sokolov AA, Erb M, Grodd W, Tatagiba MS, Frackowiak RS, neurobiology, vol. 41. San Diego: Academic; 1997. p. 489–513. Pavlova MA. Recovery of biological motion perception and net- 98. Keele SW, Ivry R. Does the cerebellum provide a common compu- work plasticity after cerebellar tumor removal. Cortex. 2014. doi:10. tation for diverse tasks? A timing hypothesis. Ann N Y Acad Sci. 1016/j.cortex.2014.05.012. 1990;608:179–207. 121. Grèzes J, Frith CD, Passingham RE. Inferring false beliefs from the 99. Pavlova MA. Biological motion processing as a hallmark of social actions of oneself and others: an fMRI study. Neuroimage. 2004;21: cognition. Cereb Cortex. 2012;22:981–95. 744–50. 100. Johansson G. Visual perception of biological motion and a model 122. Ohnishi T, Moriguchi Y, Matsuda H, Mori T, Hirakata M, for its analysis. Percept Psychophys. 1973;14:201–11. Imabayashi E, et al. The neural network for the mirror system and 101. Runeson S, Frykholm G. Kinematic specification of dynamics as an mentalizinginnormallydeveloped children:anfMRIstudy. informational basis for person and action perception: expectation, Neuroreport. 2004;15:1483–7. gender recognition, and deceptive intention. J Exp Psychol. 123. Gobbini MI, Koralek AC, Bryan RE, Montgomery KJ, Haxby JV. 1983;112:585–615. Two takes on the social brain: a comparison of theory of mind tasks. 102. Saygin AP. Superior temporal and premotor brain areas necessary J Cogn Neurosci. 2007;19:1803–14. for biological motion perception. Brain. 2007;130:2452–61. 124. Jack A, Pelphrey KA. Neural correlates of animacy attribution 103. Bonda E, Petrides M, Ostry D, Evans A. Specific involvement of include neocerebellum in healthy adults. Cereb Cortex. 2014. doi: human parietal systems and the amygdala in the perception of 10.1093/cercor/bhu146. biological motion. J Neurosci. 1996;16:3737–44. 125. Jack A, Englander ZA, Morris JP. Subcortical contributions to 104. Vaina LM, Solomon J, Chowdhury S, Sinha P, Belliveau JW. effective connectivity in brain networks supporting imitation. Functional neuroanatomy of biological motion perception in Neuropsychologia. 2011;49:3689–98. humans. Proc Natl Acad Sci U S A. 2001;98:11656–61. 126. Cullen KE. The neural encoding of self-motion. Curr Opin 105. Pavlova M, Lutzenberger W, Sokolov A, Birbaumer N. Dissociable Neurobiol. 2011;21:587–95. cortical processing of recognizable and non-recognizable biological 127. Goldberg JM, Wilson VJ, Cullen KE, et al. The cerebellum and the movement: analysing gamma MEG activity. Cereb Cortex. vestibular system. In: Goldberg JM, Wislon VJ, Cullen KE, 2004;14:181–8. Angelaki DE, Broussard DM, Buttner-Ennever J, Fukushima K, 106. Oram MW, Perrett DI. Integration of form and motion in the anterior Minor LB, editors. The vestibular system, a sixth sense. Oxford: superior temporal polysensory area (STPa) of the macaque monkey. Oxford University Press; 2012. J Neurophysiol. 1996;76:109–29. 128. Bense S, Janusch B, Vucurevic G, Bauermann T, Schlindwein P, 107. Pelphrey KA, Mitchell TV, McKeown MJ, Goldstein J, Allison T, Brandt T, et al. Brainstem and cerebellar fMRI-activation during McCarthy G. Brain activity evoked by the perception of human horizontal and vertical optokinetic stimulation. Exp Brain Res. walking: controlling for meaningful coherent motion. J Neurosci. 2006;174:312–23. 2003;23:6819–25. 129. Kleinschmidt A, Thilo KV, Büchel C, Gresty MA, Bronstein AM, 108. Pavlova MA, Sokolov AN, Bidet-Ildei C. Sex differences in the Frackowiak RS. Neural correlates of visual-motion perception as neuromagnetic cortical response to body motion. Cereb Cortex. object- or self-motion. Neuroimage. 2002;16:873–82. 2014. doi:10.1093/cercor/bhu175. 130. Bronstein AM, Grunfeld EA, Faldon M, Okada T. Reduced self- 109. Grossman E, Donnelly M, Price R, Pickens D, Morgan V, Neighbor motion perception in patients with midline cerebellar lesions. G, et al. Brain areas involved in perception of biological motion. J Neuroreport. 2008;19:691–3. Cogn Neurosci. 2000;12:711–20. 131. Bertolini G, Ramat S, Bockisch CJ, Marti S, Straumann D, Palla A. 110. Strick PL, Dum RP, Fiez JA. Cerebellum and nonmotor function. Is vestibular self-motion perception controlled by the velocity stor- Annu Rev Neurosci. 2009;32:413–34. age? Insights from patients with chronic degeneration of the 111. Glickstein M, Gerrits N, Kralj-Hans I, Mercier B, Stein J, Voogd J. vestibulo-cerebellum. PLoS One. 2012;7:6. Visual pontocerebellar projections in the macaque. J Comp Neurol. 132. Merfeld DM, Zupan L, Peterka RJ. Humans use internal models to 1994;349:51–72. estimate gravity and linear acceleration. Nature. 1999;398:615–618. 112. Schmahmann JD, Pandya DN. Course of the fiber pathways to pons 133. Laurens J, Meng H, Angelaki DE. Neural representation of orien- from parasensory association areas in the rhesus monkey. J Comp tation relative to gravity in the macaque cerebellum. Neuron. Neurol. 1992;326:159–79. 2013;80:1508–18. 218 Cerebellum (2015) 14:197–220 134. Brooks JX, Cullen KE. Multimodal integration in rostral fastigial 156. Knolle F, Schraeger E, Baess P, Kotz SA. Cerebellar contributions nucleus provides an estimate of body movement. J Neurosci. to the prediction of self-initiated sounds. Cortex. 2013;49:2449–61. 2009;29:10499–511. 157. Friston K. The free-energy principle: a unified brain theory? Nat 135. Kleine JF, Guan Y, Kipiani E, Glonti L, Hoshi M, Buttner U. Trunk Rev Neurosci. 2010;11:127–38. position influences vestibular responses of fastigial nucleus neurons 158. Blakemore SJ, Frith CD, Wolpert DM. Spatio-temporal prediction in the alert monkey. J Neurophysiol. 2004;91:2090–100. modulates the perception of self-produced stimuli. J Cogn Neurosci. 136. Shaikh AG, Meng H, Angelaki DE. Multiple reference frames for 1999;11:551–9. motion in the primate cerebellum. J Neurosci. 2004;24:4491–7. 159. Kitazawa S, Kohno T, Uka T. Effects of delayed visual information 137. Brooks JX, Cullen KE. The primate cerebellum selectively encodes on the rate and amount of prism adaptation in the human. J unexpected self-motion. Curr Biol. 2013;23:947–55. Neurosci. 1995;15:7644–52. 138. Rochefort C, Arabo A, André M, Poucet B, Save E, Rondi-Reig L. 160. Tesche CD, Karhu JJ. Anticipatory cerebellar responses during Cerebellum shapes hippocampal spatial code. Science. 2011;334: somatosensory omission in man. Hum Brain Mapp. 2000;9:119–42. 385–9. 161. Wu X, Nestrasil I, Ashe J, Tuite P, Bushara K. Inferior olive 139. Batton 3rd RR, Jayaraman A, Ruggiero D, Carpenter MB. Fastigial response to passive tactile and visual stimulation with variable efferent projections in the monkey: an autoradiographic study. J interstimulus intervals. Cerebellum. 2010;9:598–602. Comp Neurol. 1977;174:281–305. 162. Moberget T, Karns CM, Deouell LY, Lindgren M, Knight RT, Ivry 140. Kamishina H, Conte WL, Patel SS, Tai RJ, Corwin JV, Reep RL. RB. Detecting violations of sensory expectancies following cere- Cortical connections of the rat lateral posterior thalamic nucleus. bellar degeneration: a mismatch negativity study. Brain Res. 2009;1264:39–56. Neuropsychologia. 2008;46:2569–79. 141. Moulton EA, Schmahmann JD, Becerra L, Borsook D. The cere- 163. Händel B, Thier P, Haarmeier T. Visual motion perception deficits bellum and pain: passive integrator or active participator? Brain Res due to cerebellar lesions are paralleled by specific changes in Rev. 2010;65:14–27. cerebro-cortical activity. J Neurosci. 2009;29:15126–33. 142. Saab CY, Willis WD. The cerebellum: organization, functions and 164. O’Reilly JX, Mesulam MM, Nobre AC. The cerebellum predicts the its role in nociception. Brain Res Rev. 2003;42:85–95. timing of perceptual events. J Neurosci. 2008;28:2252–60. 143. Ekerot CF, Garwicz M, Schouenborg J. The postsynaptic dorsal 165. Grube M, Cooper FE, Chinnery PF, Griffiths TD. Dissociation of column pathway mediates cutaneous nociceptive information to duration-based and beat-based auditory timing in cerebellar degen- cerebellar climbing fibres in the cat. J Physiol. 1991;441:275–84. eration. Proc Natl Acad Sci U S A. 2010;107:11597–601. 144. Wu J, Chen PX. Discharge response of cerebellar Purkinje cells to 166. Koch G, Oliveri M, Caltagirone C. Neural networks engaged in stimulation of C-fiber in cat saphenous nerve. Brain Res. 1992;581: milliseconds and seconds time processing: evidence from transcra- 269–72. nial magnetic stimulation and patients with cortical or subcortical 145. Apkarian AV, Bushnell MC, Treede RD, Zubieta JK. Human brain dysfunction. Philos Trans R Soc Lond B Biol Sci. 2009;364:1907– mechanisms of pain perception and regulation in health and disease. 18. Eur J Pain. 2005;9:463–84. 167. Grube M, Lee KH, Griffiths TD, Barker AT, Woodruff PW. 146. Helmchen C, Mohr C, Erdmann C, Binkofski F. Cerebellar neural Transcranial magnetic theta-burst stimulation of the human cerebel- responses related to actively and passively applied noxious thermal lum distinguishes absolute, duration-based from relative, beat- stimulation in human subjects: a parametric fMRI study. Neurosci based perception of subsecond time intervals. Front Psychol. Lett. 2004;361:237–40. 2010;1:171. 147. Borsook D, Moulton EA, Tully S, Schmahmann JD, Becerra L. 168. Harrington DL, Lee RR, Boyd LA, Rapcsak SZ, Knight RT. Does Human cerebellar responses to brush and heat stimuli in healthy and the representation of time depend on the cerebellum? Effect of neuropathic pain subjects. Cerebellum. 2008;7:252–72. cerebellar stroke. Brain. 2004;127:561–74. 148. Moulton EA, Elman I, Pendse G, Schmahmann J, Becerra L, 169. Gooch CM, Wiener M, Wencil EB, Coslett HB. Interval timing Borsook D. Aversion-related circuitry in the cerebellum: responses disruptions in subjects with cerebellar lesions. Neuropsychologia. to noxious heat and unpleasant images. J Neurosci. 2011;31:3795– 2010;48:1022–31. 804. 170. Ivry RB, Spencer RM. Evaluating the role of the cerebellum in 149. Moulton EA, Elman I, Becerra LR, Goldstein RZ, Borsook D. The temporal processing: beware of the null hypothesis. Brain. cerebellum and addiction: insights gained from neuroimaging re- 2004;127:E13. search. Addict Biol. 2014;19:317–31. 171. Lewis PA, Miall RC. Brain activation patterns during measurement 150. Thompson RF. Neural mechanisms of classical conditioning in of sub- and supra-second intervals. Neuropsychologia. 2003;41: mammals. Philos Trans R Soc Lond B Biol Sci. 1990;329:161–70. 1583–92. 151. Koekkoek SK, Hulscher HC, Dortland BR, Hensbroek RA, 172. Bueti D, Lasaponara S, Cercignani M, Macaluso E. Learning about Elgersma Y, Ruigrok TJ, et al. Cerebellar LTD and learning- time: plastic changes and interindividual brain differences. Neuron. dependent timing of conditioned eyelid responses. Science. 2012;75:725–37. 2003;301:1736–9. 173. Wiener M, Lee YS, Lohoff FW, Coslett HB. Individual differences 152. Perrett SP, Ruiz BP, Mauk MD. Cerebellar cortex lesions disrupt in the morphometry and activation of time perception networks are learning-dependent timing of conditioned eyelid responses. J influenced by dopamine genotype. Neuroimage. 2014;89:10–22. Neurosci. 1993;13:1708–18. 174. Hayashi MJ, Kantele M, Walsh V, Carlson S, Kanai R. Dissociable 153. Bares M, Lungu O, Liu T, Waechter T, Gomez CM, Ashe J. neuroanatomical correlates of subsecond and suprasecond time Impaired predictive motor timing in patients with cerebellar disor- perception. J Cogn Neurosci. 2014;26:1685–93. ders. Exp Brain Res. 2007;180:355–65. 175. Paquette S, Mignault Goulet G, Rothermich K. Prediction, attention 154. Rahmati N, Owens CB, Bosman LW, Spanke JK, Lindeman S, and unconscious processing in hierarchical auditory perception. Gong W, et al. Cerebellar potentiation and learning a whisker- Front Psychol. 2013;4:955–6. based object localization task with a time response window. J 176. Clark A. Whatever next? Predictive brains, situated agents, and the Neurosci. 2014;34:1949–62. future of cognitive science. Behav Brain Sci. 2013;36:181–204. 155. Shadmehr R, Smith MA, Krakauer JW. Error correction, sensory 177. Popa LS, Hewitt AL, Ebner TJ. Predictive and feedback perfor- prediction, and adaptation in motor control. Annu Rev Neurosci. mance errors are signaled in the simple spike discharge of individual 2010;33:89–108. Purkinje cells. J Neurosci. 2012;32:15345–58. Cerebellum (2015) 14:197–220 219 178. Molinari M, Chiricozzi F, Clausi S, Tedesco A, De Lisa M, Leggio 200. Fatemi SH, Aldinger KA, Ashwood P, et al. Consensus paper: M. Cerebellum and detection of sequences, from perception to pathological role of the cerebellum in autism. Cerebellum. cognition. Cerebellum. 2008;7:611–5. 2012;11:777–807. 179. Braitenberg V, Heck D, Sultan F. The detection and generation of 201. Larson JCG, Mostofsky SH. Evidence that the pattern of sequences as a key to cerebellar function: experiments and theory. visuomotor sequence learning is altered in children with autism. Behav Brain Sci. 1997;20:229–77. Autism Res. 2008;1:341–53. 180. Ivry R. Exploring the role of the cerebellum in sensory anticipation 202. Remillard G. The study of sequence learning in individuals with and timing: commentary on Tesche and Karhu. Hum Brain Mapp. schizophrenia: a critical review of the literature. J Neuropsychol. 2000;9:115–8. 2014;8:231–45. 181. Nixon PD. The role of the cerebellum in preparing responses to 203. Bower JM. Is the cerebellum sensory for motor’s sake, of motor for predictable sensory events. Cerebellum. 2003;2:114–22. sensory’s sake: the view from the whiskers of a rat? Prog Brain Res. 182. Restuccia D, Della MG, Valeriani M, Leggio MG, Molinari M. 1997;114:483–516. Cerebellar damage impairs detection of somatosensory input chang- 204. Bower JM, Kassel J. Variability in tactile projection patterns to es. A somatosensory mismatch-negativity study. Brain. 2007;130: cerebellar folia crus IIa of the Norway rat. J Comp Neurol. 276–87. 1990;302:768–78. 183. Ito M. Cerebellar circuitry as a neuronal machine. Prog Neurobiol. 205. Bower JM. Computational structure of the cerebellar molecular 2006;78:272–303. layer. In: Manto M, Gruol D, Schmahmann J, Koibuchi N, Rossi 184. Molinari M, Leggio MG, Solida A, Ciorra R, Misciagna S, Silveri F, editors. Handbook of the cerebellum and cerebellar disorders. MC, et al. Cerebellum and procedural learning: evidence from focal New York: Springer; 2013. p. 1359–80. cerebellar lesions. Brain. 1997;120:1753–62. 206. Parsons LM, Bower JM, Gao JH, Xiong J, Li J, Fox PT. Lateral 185. Leggio MG, Chiricozzi FR, Clausi S, Tedesco AM, Molinari M. cerebellar hemispheres actively support sensory acquisition and The neuropsychological profile of cerebellar damage: the sequenc- discrimination rather than motor control. Learn Mem. 1997;4:49– ing hypothesis. Cortex. 2011;47:137–44. 62. 186. Dirnberger G, Novak J, Nasel C, Zehnter M. Separating coordina- 207. Liu Y, Pu Y, Gao JH, Parsons LM, Xiong J, Liotti M, et al. The tive and executive dysfunction in cerebellar patients during motor human red nucleus and lateral cerebellum in supporting roles for skill acquisition. Neuropsychologia. 2010;48:1200–8. sensory information processing. Hum Brain Mapp. 2000;10:147– 187. Doyon J, Gaudreau D, Laforce RJ, Castonguay M, Bedard PJ, 59. Bedard F, et al. Role of the striatum, cerebellum, and frontal lobes 208. Petacchi A, Laird AR, Fox PT, Bower JM. Cerebellum and auditory in the learning of a visuomotor sequence. Brain Cogn. 1997;34: function: an ALE meta-analysis of functional neuroimaging studies. 218–45. Hum Brain Mapp. 2005;25:118–228. 188. Ferrucci R, Brunoni A, Parazzini M, Vergari M, Rossi E, Fumagalli 209. Kandel ER, Schwartz JH, Jessel TM. Principles of neural science. M, et al. Modulating human procedural learning by cerebellar Norwalk: Appleton and Lange; 1991. transcranial direct current stimulation. Cerebellum. 2013;12:485– 210. Tinazzi M, Morgante F, Peretti A, Mariotti C, Panzeri M, Fiorio M, 92. et al. Impaired temporal processing of tactile and proprioceptive 189. Gomez-Beldarrain M, Garcia-Monco JC, Rubio B, Pascual-Leone stimuli in cerebellar degeneration. PLoS One. 2013;8:e78628. A. Effect of focal cerebellar lesions on procedural learning in the 211. Bhanpuri NH, Okamura AM, Bastian AJ. Active force perception serial reaction time task. Exp Brain Res. 1998;120:25–30. depends on cerebellar function. J Neurophysiol. 2012;107:1612–20. 190. Nixon PD, Passingham RE. The cerebellum and cognition: cerebel- 212. Christensen A, Giese MA, Sultan F, Mueller OM, Goericke SL, Ilg lar lesions impair sequence learning but not conditional visuomotor W, et al. An intact action–perception coupling depends on the learning in monkeys. Neuropsychologia. 2000;38:1054–72. integrity of the cerebellum. J Neurosci. 2014;34:6707–16. 191. Timmann D, Drepper J, Calabrese S, Burgerhoff K, Maschke M, 213. Ackermann H, Mathiak K, Riecker A. The contribution of the Kolb FP, et al. Use of sequence information in associative learning cerebellum to speech production and speech perception: clinical in control subjects and cerebellar patients. Cerebellum. 2004;3:75– and functional imaging data. Cerebellum. 2007;6:202–13. 82. 214. Jokisch D, Troje NF, Koch B, Schwarz M, Daum I. Differential 192. Leggio MG, Tedesco AM, Chiricozzi FR, Clausi S, Orsini A, involvement of the cerebellum in biological and coherent motion Molinari M. Cognitive sequencing impairment in patients with focal perception. Eur J Neurosci. 2005;21:3439–46. or atrophic cerebellar damage. Brain. 2008;13:1332–43. 215. Lee KH, Egleston PN, Brown WH, Gregory AN, Barker AT, 193. Gomot M, Wicker B. A challenging, unpredictable world for people Woodruff PW. The role of the cerebellum in subsecond time per- with autism spectrum disorder. Int J Psychophysiol. 2012;83:240–7. ception: evidence from repetitive transcranial magnetic stimulation. 194. Ford JM, Mathalon DH. Anticipating the future: automatic predic- J Cogn Neurosci. 2007;19:147–57. tion failures in schizophrenia. Int J Psychophysiol. 2012;83:232–9. 216. Bower JM. Functional implications of tactile projection patterns to 195. Horga G, Schatz KC, Abi-Dargham A, Peterson BS. Deficits in the lateral hemispheres of the cerebellum of the albino rat: the legacy predictive coding underlie hallucinations in schizophrenia. J of Wally Welker. Ann N Y Acad Sci. 2011;1225:130–41. Neurosci. 2014;34:8072–82. 217. Proville RD, Spolidoro M, Guyon N, Dugue GP, Selimi F, Isope P, 196. Shergill SS, White TP, Joyce DW, Bays PM, Wolpert DM, Frith CD. et al. Cerebellum involvement in cortical sensorimotor circuits for the Functional magnetic resonance imaging of impaired sensory pre- control of voluntary movements. Nat Neurosci. 2014;17:1233–9. diction in schizophrenia. JAMA Psychiatry. 2014;71:28–35. 218. Flourens P. Recherches Experimentales Sur Les Proprietes Et Les 197. Jack A, Morris JP. Neocerebellar contributions to social perception Fonctions Du Systeme Nerveux, Dans Les Animaux Vertebres. in adolescents with autism spectrum disorder. Dev Cogn Neurosci. Crevot: Paris; 1824. 2014;10:77–92. 219. Holmes G. The cerebellum of man. Brain. 1939;62:1–30. 198. Sachse M, Schlitt S, Hainz D, Ciaramidaro A, Walter H, Poustka F, 220. Manto M, Bower JM, Conforto AB, Delgado-Garcia JM, da Guarda et al. Facial emotion recognition in paranoid schizophrenia and SN, Gerwig M, et al. Consensus paper: roles of the cerebellum in autism spectrum disorder. Schizophr Res. 2014. doi:10.1016/j. motor control—the diversity of ideas on cerebellar involvement in schres.2014.08.030. movement. Cerebellum. 2012;11(2):457–87. 199. Andreasen NC, Pierson R. The role of the cerebellum in schizo- 221. Schlerf JE, Xu J, Klemfuss NM, Griffiths TL, Ivry RB. Individuals phrenia. Biol Psychiatry. 2008;64:81–8. with cerebellar degeneration show similar adaptation deficits with 220 Cerebellum (2015) 14:197–220 large and small visuomotor errors. J Neurophysiol. 2013;109:1164– 234. Kellermann T, Regenbogen C, De Vos M, Mößnang C, Finkelmeyer 73. A, Habel U. Effective connectivity of the human cerebellum during 222. Thompson RF, Steinmetz JE. The role of the cerebellum in classical visual attention. J Neurosci. 2012;32:11453–60. conditioning of discrete behavioral responses. Neuroscience. 235. Ravizza SM, McCormick CA, Schlerf JE, Justus T, Ivry RB, Fiez 2009;162:732–55. JA. Cerebellar damage produces selective deficits in verbal working 223. Huang C, Ricklefs RE. Cerebellum and human evolution: a com- memory. Brain. 2006;129:306–20. parative and information theory perspective. In: Manto M, Gruol D, 236. Robinson DA. The use of control systems analysis in the neuro- Schmahmann J, Koibuchi N, Rossi F, editors. Handbook of the physiology of eye movements. Annu Rev Neurosci. 1981;4:463– cerebellum and cerebellar disorders. New York: Springer; 2013. p. 503. 1337–57. 237. Lisberger SG. Neural basis for motor learning in the vestibuloocular 224. Kawashima R, Okuda J, Umetsu A, Sugiura M, Inoue K, Suzuki K, reflex of primates. III. Computational and behavioral analysis of the et al. Human cerebellum plays an important role in memory-timed sites of learning. J Neurophysiol. 1994;72:974–98. finger movement: an fMRI study. J Neurophysiol. 2000;83:1079–87. 238. Westheimer G, McKee SP. Visual acuity in the presence of retinal- 225. D’Angelo E, Casali S. Seeking a unified framework for cerebellar image motion. J Opt Soc Am. 1975;65:847–50. function and dysfunction: from circuit operations to cognition. Front 239. Little WJ. On the influence of abnormal parturition, difficult la- Neural Circ. 2012;6:116. bours, premature birth and asphxia neonatorum on the mental and 226. Schlerf J, Ivry RB, Diedrichsen J. Encoding of sensory physical condition of the child, especially in relation to deformities. prediction errors in the human cerebellum. J Neurosci. Clin Orthop Relat Res. 1966;46:7–22. 2012;32:4913–22. 240. Spence RD, Zhen Y, White S, Schlinger BA, Day LB. Recovery of 227. Ivry RB, Spencer RMC. The neural representation of time. Curr motor and cognitive function after cerebellar lesions in a songbird: Opin Neurobiol. 2004;14:225–32. role of estrogens. Eur J Neurosci. 2009;29:1225–34. 228. Holscher C, Munk M. Information processing by neuronal popula- 241. Williams AJ, Ling GSF, Tortella FC. Severity level and injury track tions. Cambridge: Cambridge University Press; 2008. determine outcome following a penetrating ballistic-like brain inju- 229. Frings M, Boenisch R, Gerwig M, Diener HC, Timmann D. ry in the rat. Neurosci Lett. 2006;408:183–8. Learning of sensory sequences in cerebellar patients. Learn Mem. 242. Willson ML, Bower AJ, Sherrard RM. Developmental neural plas- 2004;11:347–55. ticity and its cognitive benefits: olivocerebellar reinnervation com- 230. Pfordresher PQ, Mantell JT, Brown S, Zivadinov R, Cox JL. pensates for spatial function in the cerebellum. Eur J Neurosci. Brain responses to altered auditory feedback during musical 2007;25:1475–83. keyboard production: an fMRI study. Brain Res. 2014;1556: 243. Konczak J, Schoch B, Dimitrova A, Gizewski E, Timmann D. 28–37. Functional recovery of children and adolescents after cerebellar 231. Bell C, Bodznick D, Montgomery J, Bastian J. The generation and tumour resection. Brain. 2005;128:1428–41. subtraction of sensory expectations within cerebellum-like struc- 244. Fatemi SH. Cerebellum and autism. Cerebellum. 2013;12:778–9. tures. Brain Behav Evol. 1997;50:17–31. 245. Iwanaga R, Honda S, Nakane H, Tanaka K, Toeda H, Tanaka G. 232. Dancey E, Murphy B, Srbely J, Yielder P. The effect of experimental Pilot study: efficacy of sensory integration therapy for Japanese pain on motor training performance and sensorimotor integration. children with high-functioning autism spectrum disorder. Occup Exp Brain Res. 2014;232:2879–89. Ther Int. 2014;21:4–11. 233. Gottwald B, Mihajlovic Z, Wilde B, Mehdorn HM. Does the 246. Pavlova M, Krageloh-Mann I, Sokolov A, Birbaumer N. cerebellum contribute to specific aspects of attention? Recognition of point-light biological motion displays by Neuropsychologia. 2003;41:1452–60. young children. Perception. 2001;30:925–33. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Cerebellum Unpaywall

Consensus Paper: The Role of the Cerebellum in Perceptual Processes

Download PDF
24 pages

Loading...

Page 2

Loading...

Page 3

Loading...

Page 4

Loading...

Page 5

Loading...

Page 6

Loading...

Page 7

Loading...

Page 8

Loading...

Page 9

Loading...

Page 10

Loading...

Page 11

Loading...

Page 12

Loading...

Page 13

Loading...

Page 14

Loading...

Page 15

Loading...

Page 16

Loading...

Page 17

Loading...

Page 18

Loading...

Page 19

Loading...

Page 20

Loading...

Page 21

Loading...

Page 22

Loading...

Page 23

Loading...

Page 24

 
/lp/unpaywall/consensus-paper-the-role-of-the-cerebellum-in-perceptual-processes-zLHBNSdsdn

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
1473-4222
DOI
10.1007/s12311-014-0627-7
Publisher site
See Article on Publisher Site

Abstract

Cerebellum (2015) 14:197–220 DOI 10.1007/s12311-014-0627-7 CONSENSUS PAPER Consensus Paper: The Role of the Cerebellum in Perceptual Processes Oliver Baumann & Ronald J. Borra & James M. Bower & Kathleen E. Cullen & Christophe Habas & Richard B. Ivry & Maria Leggio & Jason B. Mattingley & Marco Molinari & Eric A. Moulton & Michael G. Paulin & Marina A. Pavlova & Jeremy D. Schmahmann & Arseny A. Sokolov Published online: 6 December 2014 The Author(s) 2014. This article is published with open access at Springerlink.com Abstract Various lines of evidence accumulated over the past predictive processing, and perceptual sequencing. While no 30 years indicate that the cerebellum, long recognized as single explanation has yet emerged concerning the role of the essential for motor control, also has considerable influence cerebellum in perceptual processes, this consensus paper sum- on perceptual processes. In this paper, we bring together marizes the impressive empirical evidence on this problem experts from psychology and neuroscience, with the aim of and highlights diversities as well as commonalities between providing a succinct but comprehensive overview of key existing hypotheses. In addition to work with healthy individ- findings related to the involvement of the cerebellum in sen- uals and patients with cerebellar disorders, it is also apparent sory perception. The contributions cover such topics as ana- that several neurological conditions in which perceptual dis- tomical and functional connectivity, evolutionary and com- turbances occur, including autism and schizophrenia, are as- parative perspectives, visual and auditory processing, biolog- sociated with cerebellar pathology. A better understanding of ical motion perception, nociception, self-motion, timing, the involvement of the cerebellum in perceptual processes will O. Baumann (*) J. B. Mattingley M. Leggio Queensland Brain Institute, The University of Queensland, St. Lucia, Department of Psychology, Sapienza University of Rome, Rome, Queensland, Australia Italy e-mail: o.baumann@uq.edu.au M. Leggio M. Molinari R. J. Borra I.R.C.C.S. Santa Lucia Foundation, Rome, Italy Department of Radiology and Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard E. A. Moulton Medical School, Charlestown, MA, USA Pain/Analgesia Imaging Neuroscience (P.A.I.N.) Group, Department of Anesthesia, Boston Children’s Hospital, Center for Pain and the R. J. Borra Brain, Harvard Medical School, Waltham, MA, USA Department of Diagnostic Radiology, Medical Imaging Centre of Southwest Finland, Turku University Hospital, Turku, Finland M. G. Paulin Department of Zoology, University of Otago, Otago, New Zealand J. M. Bower Numedon Inc., Pasadena, CA, USA M. A. Pavlova Department of Biomedical Magnetic Resonance, Medical School, K. E. Cullen Eberhard Karls University of Tübingen, Tübingen, Germany Department of Physiology, McGill University Montreal, Montreal, Canada J. D. Schmahmann Ataxia Unit, Cognitive Behavioral Neurology Unit, Laboratory for C. Habas Neuroanatomy and Cerebellar Neurobiology Department of Service de NeuroImagerie, CHNO des Quinze-Vingts, UPMC Paris Neurology, Massachusetts General Hospital and Harvard Medical 6, Paris, France School, Boston, MA, USA R. B. Ivry A. A. Sokolov Department of Psychology, University of California, Berkeley, CA, Département des Neurosciences Cliniques, Centre Hospitalier USA Universitaire Vaudois (CHUV), Lausanne, Switzerland 198 Cerebellum (2015) 14:197–220 thus likely be important for identifying and treating perceptual cerebellum in pain perception is reviewed by Drs. Borra and deficits that may at present go unnoticed and untreated. This Moulton. Dr. Ivry presents a hypothesis and data to suggest paper provides a useful framework for further debate and that the cerebellum acts as a timing device for motor and non- empirical investigations into the influence of the cerebellum motor processes. Drs. Leggio and Molinari present evidence on sensory perception. for a model that posits a central role for the cerebellum in the detection and prediction of perceptual sequences. The review . . . Keywords Audition Biological motion Cerebellum closes with a contribution from Dr. Bower, who suggests that . . . . . Connectivity Evolution fMRI Pain Perception the cerebellum is not itself involved in perceptual processing, . . . . Prediction Single-unitrecording Self-motion Sequencing but instead, its influence on perception as well as motor . . State estimation Timing Vision control, is indirect through its role in monitoring and adjusting the acquisition of sensory data. Introduction Anatomical Circuits Relevant to the Role For 150 years, functional analyses of the cerebellum have of the Cerebellum in Perception (J.D. Schmahmann) focused on the role of this subcortical structure in the control and coordination of movement. In the past 30 years, however, The cerebellar role in perception is predicated on the fact that clinical, experimental, and neuroimaging studies have provided it is an essential node in the distributed neural circuits compelling evidence for the involvement of the cerebellum in subserving sensorimotor, autonomic, and cognitive function task domains as diverse as memory, language, and emotion. as well as emotional processing. The following is a summary Crucially, several lines of evidence suggest that the cerebellum of these pathways and connections. For earlier comprehensive has an influence on perceptual functions. Observations from reviews and original citations, please see Schmahmann [1–3] anatomical and electrophysiological studies in monkeys and and Schmahmann and Pandya [4]. cats indicate the existence of cerebellar connections with visual- and auditory-related cortices. Moreover, clinical reports Peripheral Afferents in humans have revealed that both focal and diffuse lesions of the cerebellum lead to a wide range of sensory impairments. Auditory and visual inputs are conveyed from primary sensory While damage to the cerebellum does not cause a complete loss receptors to vermal lobules VI and VII [5], and visual inputs of sensory function, it is apparent that several sensory and also reach the dorsal paraflocculus. Spinocerebellar tracts ter- perceptual processes are affected, such as motion and time minate in the sensorimotor cerebellum in the anterior lobe and perception, or the ability to recognize perceptual sequences. lobule VIII [6], while vestibular afferents target lobule X [7]. In this consensus paper, we summarize key findings and Climbing fibers from the sensorimotor-recipient inferior olivary concepts with the aim of demonstrating and explaining the nuclei project to the sensorimotor cerebellum; the principal cerebellar influence on perceptual tasks. To this end, we have olivary nucleus is devoid of peripheral inputs and is linked with gathered contributions from 14 experts in complementary the cognitive cerebellum in the posterior lobe (see [3]). fields of psychology and neuroscience, providing a range of different and sometimes controversial viewpoints. We believe Cerebrocerebellar Pathways that a new consensus that draws on and integrates the ideas presented here will help unravel the enigmatic role or influ- Cerebellar connections with the cerebral cortex include two- ence of the cerebellum in perceptual processing. The review stage feedforward and feedback loops with obligatory synap- begins with a succinct overview of the anatomical connections ses in the pons and thalamus. The top-down circuit is of the cerebellum with sensory and perceptual areas in the corticopontine–pontocerebellar and the bottom-up is cerebrum by Dr. Schmahmann. Dr. Habas then provides an cerebellothalamic–thalamocortical. evaluation of the functional connections between the cerebel- lum and cerebral perceptual systems, drawing on studies using Corticopontine Projections modern neuroimaging techniques. Dr. Paulin provides an evolutionary and comparative perspective on cerebellar in- Knowledge of the corticopontine projections provides critical volvement in perceptual functions. Evidence for a cerebellar insights into the nature of the information to which the cere- role in visual and auditory processing is summarized by Drs. bellum has access. Projections arise from neurons in layer Vb Baumann and Mattingley, followed by a commentary from of sensorimotor regions in the precentral, premotor, and sup- Drs. Pavlova and Sokolov on visual processing of biological plementary motor area, primary somatosensory cortices, and motion. Dr. Cullen writes on the critical function of the cere- the rostral parietal lobe [8–11]. Studies in stroke patients also bellum in self-motion perception. Evidence for a role of the show topography of motor function in the human pons [12]. Cerebellum (2015) 14:197–220 199 Considerable corticopontine projections are derived also behavioral manifestations. The superior parietal lobule con- from the prefrontal cortex, multimodal regions of the posterior cerned with multiple joint position sense, touch, and proprio- parietal and temporal lobes, paralimbic cortices in the cingu- ceptive impulses projects throughout central and lateral re- late and posterior parahippocampal gyrus, and visual associ- gions of the rostrocaudal pons. The caudal inferior parietal ation cortices in the parastriate region, supporting multimodal, lobule implicated in the neglect syndrome favors the rostral supramodal, and limbic related functions necessary for per- half of the pons in the lateral and dorsolateral regions [10]. ception (Fig. 1). Auditory association areas in the superior temporal gyrus Prefrontopontine projections arise from dorsolateral and and supratemporal plane are connected with the lateral and dorsomedial convexities concerned with attention and conju- dorsolateral pontine nuclei. Cortices in the upper bank of the gate eye movements (area 8), spatial attributes of memory and superior temporal sulcus activated during face recognition working memory (area 9/46d), planning, foresight, and judg- tasks project to the lateral, dorsolateral, and extreme dorsolat- ment (area 10), motivational behavior and decision-making eral pontine nuclei [14]. Motion-sensitive temporal lobe areas capabilities (areas 9 and 32), and from areas 44 and 45 MT (middle temporal), FST (fundus of the superior temporal homologous to language areas in human [13]. sulcus), and MST (medial superior temporal) also have pon- Posterior parietal association cortices are critical for direct- tine connections [15], but inferotemporal cortex including the ed attention, visual–spatial analysis, and vigilance in the con- rostral lower bank of the superior temporal sulcus, which is tralateral hemispace; lesions are associated with complex relevant for feature discrimination, has no pontine efferents. Fig. 1 Composite color-coded summary diagram illustrating the I through IX are taken. c Patterns of termination within the nuclei of the distribution within the basis pontis of rhesus monkey of projections basis pontis. Other cerebral areas known to project to the pons are derived from association and paralimbic cortices in the prefrontal depicted in white. Cortical areas with no pontine projections are shown (purple), posterior parietal (blue), superior temporal (red), parastriate, in yellow (from anterograde and retrograde studies) or gray (from and parahippocampal regions (orange), and from motor, premotor and retrograde studies). Dashed lines in the hemisphere diagrams represent supplementary motor areas (green). a Medial, lateral, and orbital views of sulcal cortices. Dashed lines in the pons diagrams represent pontine the cerebral hemisphere from which the projections are derived. b Plane nuclei; solid lines depict corticofugal fibers (from [1] and [13]) of section through the pons from which the rostrocaudal levels of the pons 200 Cerebellum (2015) 14:197–220 Thus, the dorsal visual (where) stream concerned with motion and converging pontocerebellar projections led to the sugges- analysis and visual–spatial attributes of motion participates in tion that information from one cerebral cortical area is distrib- the cerebrocerebellar interaction, but the ventral visual (what) uted to numerous sites in the cerebellar cortex [27], although stream governing visual object identification does not. trans-synaptic viral tract tracing studies reveal that anterograde Parastriate projections from occipitotemporal and occipitoparietal projections through the medial pons are directed to focal areas regions also respect the dorsal–ventral dichotomy. The medial in crus I and crus II [28]. and dorsal prelunate regions project to the pons (dorsolateral, lateral, and lateral aspect of the peripeduncular nuclei most Cerebellar Feedback heavily), but ventral prelunate cortices and inferotemporal re- gions do not [16]. Projections from the temporal lobe homologue Purkinje cells convey the output of the cerebellar cortex to the of the Wernicke language area in human, together with those deep cerebellar nuclei (DCN), which send projections back to from the monkey homologue of Broca’s area, are relevant in the the brainstem, or to the cerebral cortex via the thalamus. The light of cerebellar activation during functional neuroimaging cerebellar cortex–DCN–thalamus–cerebral cortex feedback loop studies of language [17, 18] and in disorders of language follow- is arranged so that motor related interpositus nuclei (globose and ing cerebellar lesions [19, 20]. emboliform in human) send efferents from cerebellar anterior Paralimbic projections arise from posterior parahippocampal lobe motor areas to the cerebral sensorimotor regions, whereas gyrus important for spatial attributes of memory, directed to the ventral dentate sends information from the cerebellar poste- lateral, dorsolateral, and lateral peripeduncular nuclei. Cingulate rior lobe to cerebral association areas—prefrontal, posterior pa- cortex projections arise from motor areas in the depth of the rietal, and others [28, 29](seeFig. 2). The cerebellar vermis and cingulate sulcus [21] and from areas concerned with motivation fastigial nucleus are linked with brainstem and thalamic struc- and drive in rostral and caudal cingulate areas [22]. The anterior tures concerned not only with vestibular and oculomotor control, insular cortex, important for autonomic systems and pain mod- posture, and equilibrium, but also with autonomic and paralimbic ulation also has pontine connections [9]. Projections arise also cerebral areas, consistent with the notion of the vermis and from multimodal deep layers of the superior colliculus and fastigial nucleus as the limbic cerebellum [3]. medial mammillary bodies involved in memory and emotion [23]. The hypothalamus, critical for autonomic control and lim- Synthesis bic behaviors, has direct reciprocal connections with the cerebel- lum [24]. Against the backdrop of the heterogenous and topographically Corticopontine projections are arranged with topographic arranged connections of the cerebellum with the rest of the specificity. Sensorimotor terminations are more caudally situated; neuraxis stands the essentially constant architecture of the association areas project more rostrally. Terminations occur in cerebellar cortex. This dichotomy is the basis of the dysmetria multiple patches forming interdigitating mosaics. The signifi- of thought theory, which poses that a constant computation— cance of associative corticopontine inputs in human compared the universal cerebellar transform—is applied to multiple with monkey is underscored by enlargement in human of the domains of neurological function subserved by the distributed medial part of the cerebral peduncle conveying prefontopontine neural circuits of which cerebellum is an integral node [3]. The fibers [25], reflecting evolutionary pressure in which intercon- anatomical connections that link the cerebellum with both the nected systems evolve in concert with each other. external and the internal worlds thus provide the critical neural substrates of the putative cerebellar role in perception. These Pontocerebellar Projections conclusions from tract tracing studies in the monkey are supported by resting state functional connectivity magnetic The caudal pons sends sensorimotor-related information to the resonance imaging (MRI; [30]) and task-based functional cerebellar anterior lobe. Rostral pontine nuclei convey cogni- MRI studies in humans [18], as well as by clinical investiga- tively relevant information to the posterior cerebellum: medial tions in patients with cerebellar damage [19]. pontine projections from prefrontal cortices to crus I and to crus II, and medial, ventral, and lateral pons conveying infor- mation from parietal association cortices to crus I, crus II, and lobule VIIB. These anatomical studies extend earlier physio- Resting-State Functional Connectivity logical conclusions that parietal and prefrontal cortices are Between Cerebellum and Sensory Systems (C. Habas) functionally related mainly to crusI,crusII, andthe paramedian lobule of the cerebellum [26]. In the pontocerebellar projection, Measurement of human brain resting-state activity with MRI each cerebellar folium receives input from a unique comple- has allowed us to precisely determine the functional connec- ment of pontine cell groups, some of which are widely separat- tivity (FC) between specific zones of the cerebellum and the ed [1, 27]. The pattern of diverging corticopontine projections rest of the brain. FC is based on temporal correlations between Cerebellum (2015) 14:197–220 201 Fig. 2 a Diagram of the lateral view of a cebus monkey brain (top) to show the location of injections of McIntyre-B strain of herpes simplex virus I in the primary motor cortex arm representation (M1arm), ventral premotor cortex arm representation (PMVarm), and in the prefrontal cortex in areas 9 and 46. The resulting retrogradely labeled neurons (below)in the cerebellar interpositus nucleus (IP) and dentate nucleus (DN) are indicated by solid dots and show the dorsal–ventral dichotomy in dentate projections to motor versus prefrontal cortices. Adapted from [29]. b Representation on flattened views of the cerebellum of the input– output organization of cerebellar loops with motor cortex M1 (left) and area 46 (right) revealed using anterograde and retrograde strains of rabies virus as tract tracer. M1 is interconnected with lobules IV to VI; prefrontal cortical area 46 is linked predominantly with crus II. Adapted from [28] spontaneous, low-frequency (0.01–0.1 Hz) fluctuations of the lobules IV–VI, the second in lobules VIIb–VIII, and a third blood-oxygen-level-dependent (BOLD) signal at rest between in lobules VI–VIIA [36]. functionally and anatomically linked cerebral areas [31]. Two Discrepant results, however, were obtained for the visual main statistical methods are used to compute resting-state and auditory cerebellum. O’Reilly and colleagues [34]found functional maps passing through the cerebellum [32]: (1) functional coherence between visual area MT and superior independent component analysis, which is used to identify temporal gyrus, including auditory primary and associative multiple temporally cohesive, spatially distributed networks zones, with cerebellar lobules V-VI-VIII and lobules V–VI, and (2) regression analysis of activity in a region of interest respectively. Buckner and colleagues, however, failed to de- against that of the remainder of the brain. These methods tect any functional connectivity between auditory cortex and have contributed to distinguish two anatomo-functional cerebellum [30, 33]. The proximity between the occipital lobe parts of the cerebellum [33–35]: a sensorimotor region and the underlying cerebellar cortex has been proposed as a (lobules V–VI and VIII) and a prominent multimodal cog- possible explanation of the discrepancy between these data. nitive and limbic region (lobule VIIA, especially crus I and However, Sokolov et al. [37] (see also the section by Drs. II, with adjoining parts of lobule VI and VIIB, and lobule Sokolov and Pavlova, “Cerebellar Involvement in Biological IX). The sensorimotor cerebellum corresponds predomi- Motion Processing (A.A. Sokolov and M.A. Pavlova)”) nantly to sensory parts of its multiple somatotopic maps found, using DTI, structural interconnection between cerebel- that receive exteroceptive and proprioceptive inputs from lar crus I and right superior temporal sulcus (STS), in agree- spinal, trigeminal, and somatosensory cortical afferents, and ment with a previous seed-based functional connectivity result send outputs to motor areas in order to control, guide, and which showed functional coherence between STS and cere- correct ongoing movements. At least three somatotopic bellar lobules VI/VIIA [38]. It is noteworthy that no functional representations have been reliably described: the first in link was found in these two studies between cerebellum and 202 Cerebellum (2015) 14:197–220 primary visual cortex (BA 17) in line with previous animal these animals, the cerebellum is evidently involved in tracking studies. Notwithstanding, using cerebellar seed-based function- objects using the electric sense [46–48]. But comparative ana- al connectivity, Sang and colleagues [39] found correlations tomical, physiological, and behavioral evidence indicated that between visual networks and hemispheric lobules I–VI and this is not an anomaly. Across all vertebrates, the cerebellum vermal lobules VIIb–IX, as well as auditory networks and seems to have a primary role in motion analysis and motion hemispheric lobules VI-VIIb-VIIIa. Ding et al. [40] also iden- prediction, with a role in motor control a consequence of this tified decreased functional connectivity between visual cortex perceptual capability, analogous to the role of dynamical state (BA 17) and cerebellum (crus I and II, vermis of lobules VI–VII estimators in artificial control systems [49]. and tonsilae) when they compared ambliopic patients with The theory that cerebellum is a neural analogue of a dy- healthy subjects. One possibility would be that amblyoply first namical state estimator simplifies and generalizes the theory induced diminished connectivity between primary visual cortex that cerebellum is engaged in motor control. An animal needs and interconnected parietal (BA 40) and prefrontal (BA 6/8) to determine the kinematic state of its own body in order to cortices, and that this altered connectivity indirectly affected the control movements, and to perceive and dynamically interact cerebellum via the prefronto-parieto-pontine pathway. with other objects and organisms. In particular, active sensing The cerebellum is also involved in the limbic ‘salience and exploratory behavior is critically dependent on accurate network,’ mainly encompassing insula, frontal operculum, information about the configuration and motion of sense medial prefrontal cortex, and hypothalamus [35], and in organs during sensory acquisition [47]. It has been shown in charge of interoceptive and autonomic processing [41]. There- human and animal studies that the cerebellum plays a crucial fore, it could be hypothesized that cerebellar zones belonging role in active sensory acquisition [50, 51]. Other tasks that to the salience network (lobules VI, VIIA, and VIIB) process have been shown to involve the cerebellum in humans also interoceptive data. Paravermal and vermal lobule VI may seem to require dynamical state estimation [52–59]. constitute a specific node receiving exteroceptive and intero- The cerebellum is a characteristic of vertebrates, but ceph- ceptive data, since it has been found active during emotional alopod molluscs (squid and octopus) appear to have evolved a responses such as disgust [42]. In conclusion, functional con- cerebellum independently. The cephalopod cerebellum re- nectivity mainly confirms previous results acquired with his- ceives visual and vestibular sense data and is involved in tological tracking and electrical stimulation and adds some whole-body and oculomotor stabilization during locomotion new insights: the ‘sensory’ cerebellum is mainly part of the [60–62]. Cephalopods are the only agile predators among sensorimotor (and vestibular) cerebellum and may also com- molluscs. prise areas that process visual, auditory, and interoceptive Cerebellar-like structures occur in a number of animal signals. Finally, there may be two distinct roles for the cere- phyla. These are distinguished from the ‘cerebellum proper’ bellum in perceptual tasks. The first involves the ‘sensory’ by a lack of climbing fibers and a lack of direct projections to cerebellum for perceptual analysis, cancellation, and anticipa- motor and premotor structures. The most well-known cere- tion based on internal models during, for instance, fine ex- bellar-like structures are electrosensory and lateral-line ploratory movements. The second involves the polymodal mechanosensory nuclei in fishes [63, 64], but they are found ‘executive’ cerebellum, which is associated with working in many vertebrates including humans [65]. They are involved memory, attention, and decision-making processes for con- in removing distortions from external signal sources caused scious elaboration of the mental representation of a perceived by an animal’s own activity. Thus, in electroreception, the object [43]. cerebellum is involved in sensing external targets by exploiting distortions in signals generated by the animal’s own activity, while cerebellar-like circuits are involved in Evolutionary Perspectives on Cerebellar Function (M.G. sensing external targets by eliminating distortions of target Paulin) signals caused by the animal’s own activity. Cerebellar-like circuits have been reported among arthro- Early in the twentieth century, studies of brain-damaged sol- pods, onychophorans, and polychaete annelids. These inver- diers led to a consensus that cerebellum is dedicated to motor tebrates are all active foragers, with appendages that support control, because focal cerebellar ablation led to obvious motor arrays of sensilla [66]. Cerebellar-like structures in insects deficits without obvious perceptual deficits [44]. Late in the may be involved in orientation and navigation [67]. They twentieth century, human functional imaging studies revealed seem to be more prominent in species like honeybees, which that the cerebellum is actively engaged in a variety of cogni- use their antennae as active probes, than in moths whose tive, perceptual, and behavioral tasks, even when subjects are antennae are passive receivers [66]. not moving [45]. The cerebellar cortical circuit common to the cerebellum In the middle of the twentieth century, the gigantocerebellum and cerebellar-like circuits has apparently evolved indepen- of weakly electric fish stood out as an anomaly because, in dently in at least five groups of animals: vertebrates, Cerebellum (2015) 14:197–220 203 cephalopod molluscs, arthropods, onychophorans, and poly- predominantly in vermal lobule VII and hemispheric lobules chaete annelids. All species in which cerebellar and/or VI, that were differentially activated for visual stimuli and cerebellar-like circuits have been reported are motile and auditory stimuli. In the 1980s, several laboratories started to sufficiently large that their kinematics is influenced by inertia, use neuronal tracers to examine cerebrocerebellar projections and they interact with other such animals. Inertia constrains in non-human primates and discovered that visual as well as how the kinematic state (position, configuration, and rates of auditory association areas are anatomically connected with the change) of an object changes as a function of applied force, cerebellum [2] (see also the section by Dr. Schmahmann, such that, if an object has inertia, then information about its “Anatomical Circuits Relevant to the Role of the Cerebellum kinematic state can be used to predict its future position and in Perception (J.D. Schmahmann)”). Interestingly, while cer- configuration at least in the short term. This is not true of ebellar connections were found for dorsal visual stream areas, animals (or indeed objects of any kind) whose mass is small or which are known to underlie motion analysis, this was not the drag is large relative to applied forces [68]. Animals that have case for ventral visual stream areas, which are involved in evolved cerebellar(-like) circuits are, therefore, animals for visual object recognition. This finding suggests that the cere- which probabilistic inference about the kinematic states of bellum is particularly involved in processing dynamic (i.e., self and others is both possible and useful. The fact that this time varying) visual information. group includes disparate, unrelated species indicates that the The first evidence in humans for a cerebellar involvement genetic and developmental capacity for cerebellar(-like) cir- in visual processes derives from work undertaken by Ivry and cuits may be shared by all animals with nervous systems and Diener, who found that cerebellar patients were impaired in that it has been co-opted by evolution whenever there has been making judgments of the velocity of moving stimuli, whereas an ecological opportunity for animals capable of dynamic elementary visual functions remained intact [86]. These find- motion prediction and control [69]. More generally, the ability ings were later corroborated and extended by Thier and to predict state trajectories of dynamical systems from obser- Haarmeier, who reported that patients with cerebellar lesions vations provides a core capability that may underpin a wide were also impaired in detecting and discriminating moving variety of perceptual, cognitive, and motor tasks [70]. visual signals in the presence of visual noise [87]. Similarly, it Until a few years ago, the Kalman filter was the only was found that cerebellar lesions can disturb auditory process- known practical algorithm for dynamical state estimation ing, by significantly increasing thresholds in duration [88]and [71]. It assumes linear target dynamics, an assumption that pitch discrimination tasks [57]. does not hold for mechanical linkages like human and animal Despite evidence of a sensory processing role for the cere- bodies. Newer algorithms based on drawing random samples bellum, the exact manner in which visual and auditory informa- from probability distributions defined by observations are able tion is represented in the human cerebellum remains unclear. To to track states of high-dimensional nonlinear systems [72]. address this issue, we used functional magnetic resonance im- These algorithms can be implemented using spiking neurons, aging (fMRI) to monitor neural activity within the cerebellum in which a spike at a particular location in a network represents while participants were engaged in a task that required them to a sample at a particular location in the state space of the system determine the direction of a visual or auditory motion signal in tracked by the network [73, 74]. There is growing evidence noise [89]. In the visual motion task, vermal lobule VI and right- that neurons use Bayesian Monte-Carlo algorithms of this hemispheric lobule X were active (see Fig. 3a), whereas in the kind to implement decisions and actions [75–83]. auditory motion task, activity was elevated in hemispheric lobules VI and VIII (see Fig. 3b). Interestingly, for both auditory and visual motion tasks, activity within left crus I increased as The Role of the Cerebellum in Visual and Auditory the strength of the motion signal decreased (see Fig. 3c), sug- Processing (O. Baumann and J.B. Mattingley) gesting that the recruitment of the cerebellum is related to the perceptual demands of a task. These findings are consistent with Over the last decade, hypotheses of human cerebellar function results from a positron emission tomography study in which have undergone dramatic revisions [84]. Of these, perhaps the similar regions of cerebellar cortex became more active as the most intriguing is the proposal that the cerebellum plays a role level of difficulty of a pitch discrimination task increased [90]. in sensory processes. In the following, we review evidence for In addition, recent neuropsychological and neuroimaging stud- cerebellar involvement in visual and auditory perception. ies have implicated left crus I in tasks involving biological Cerebellar responses to auditory and visual stimulation motion perception [91, 92] (see also section by Drs. Sokolov were described in the 1940s. Snider and Stowell [85]recorded and Pavlova, “Cerebellar Involvement in Biological Motion electrical responses in the cerebellar cortex of 150 anesthe- Processing (A.A. Sokolov and M.A. Pavlova)”), suggesting a tized cats, evoked by acoustic clicks as well low-intensity light role in higher-level visual processing. flashes. Using this approach, they revealed the existence of Interestingly, there have also been incidental reports of cere- distinct, but partially overlapping cerebellar regions, bellar activity during tasks involving crossmodal matching 204 Cerebellum (2015) 14:197–220 Fig. 3 MR brain slices showing distinct set of cerebellar regions that motion condition; green shading represents activity for the auditory were differentially activated for: a visual stimuli and b auditory stimuli, as motion condition; yellow shading indicates activation overlap between well as c showing a negative linear relationship between fMRI signal and the visual and auditory conditions). Figure reproduced with permission motion signal strength (red shading represents activity for the visual from [89] [93–95]. For example, we observed that combined audiovisual Cerebellum and Perception: The Role of the Cerebellum in motion detection led to increased activity bilaterally in cerebellar Self-Motion Perception (K.E. Cullen)”) is essential for a wide lobule VI and right lateral crus I, relative to unimodal visual and range of daily life activities such as safe car driving, motor auditory motion tasks [93]. This is consistent with findings in learning, imitation, social interaction, and non-verbal commu- monkeys that different sensory areas of the cerebral cortex con- nication through body language [99]. Healthy adults and verge on common areas within the neocerebellum [1]. Taken children easily recognize personality traits through actions of together, these results suggest that the cerebellar hemispheres others, even if they are represented through a set of light dots play a role in the detection of intermodal invariant temporal– placed on the main body joints, in “point-light biological spatial features in concurrent streams of audio-visual information. motion displays” [100, 101](see Fig. 4a). Neurophysiological A prominent hypothesis is that the cerebellum aids infor- and lesional research has revealed the core components of the mation processing by making predictions, in the form of an cortical system underlying visual perception of body motion “internal model” of sensory events [96]. An alternative ac- that includes areas in the frontal [102] and parietal [103–105] count is that the cerebellum facilitates perception by monitor- cortices, the fusiform gyrus and superior temporal sulcus ing and coordinating the acquisition of sensory information (STS) [105–107], mainly in the right brain hemisphere [97] (see the section by Dr. Bower, “Is the Cerebellum Sen- [108]. Yet, our knowledge on engagement of brain structures sory for Motor’s Sake, or Motor for Sensory’s Sake? (J.M. outside the cerebral cortex is still rather limited. Bower)”). A third hypothesis is that the cerebellum functions as Early positron emission tomography (PET) data suggest an internal timing device for both motor and perceptual process- activation of the amygdala and left lateral cerebellum for es, with different regions of the cerebellum thought to provide point-light dance-like biological motion [103]. fMRI also separate timing computations for different tasks [98](seethe indicates cerebellar activity during visual processing of body sectionbyDr. Ivry, “Sensory Processing and the Cerebellum: motion. However, the outcome is controversial, in particular, Timing (R.B. Ivry)”). At present, there is no unequivocal support in respect to topography and lateralization. Right midline for any one of these models, and in fact, each can provide at least cerebellar response was found for a contrast of canonical a partial account for many of the relevant findings. against scrambled point-light actions when observers per- In conclusion, while there is considerable evidence that the formed a one-back repetition task [109]. In a two-alternative cerebellum contributes to auditory and visual sensory process- forced choice (2AFC) discrimination task, bilateral activation es, its precise role is not yet well understood. We need more in the cerebellar hemispheres was shown for canonical and information about how the cerebellum interacts with visual scrambled point-light displays pooled together and contrasted and auditory networks, particularly in terms of the nature against baseline, with specific activation of the left lateral (inhibitory or excitatory) and directionality (feedback or cerebellar region QuP (posterior quadrangular lobule or lobule feedforward) of these connections. VI) when judging direction of biological motion [104]. Psychophysical data in patients with tumors to the left cerebellum showed that damage to the lateral lobules VIIB, Cerebellar Involvement in Biological Motion Processing VIIIA, and crus I and II substantially affects visual sensitivity (A.A. Sokolov and M.A. Pavlova) to biological motion simultaneously camouflaged by addition- al moving dots (a spatially scrambled display containing the Visual perception of bodily movements of others (for percep- same characteristics as a canonical biological motion display tion of self-motion, see the section by Dr Cullen, “The (except for the spatial positions of the dots) served as a control Cerebellum (2015) 14:197–220 205 Fig. 4 Loop between the cerebellum and superior temporal sulcus (STS) Elsevier Inc., with permission of the publisher, Elsevier. c Three- subserving biological motion perception. a Example of a point-light dimensional representation of the structural loop pathway between the biological locomotion stimulus with 11 dots placed on the main joints right STS and crus I, as revealed by diffusion tensor imaging (DTI). of the walking human body. Outline added for illustrative purpose. From Fibers descending from the STS to the cerebellum pass through the pons [246] Pion Ltd., London, www.envplan.com. b Dynamic causal and the middle cerebellar peduncle (MCP), while ascending fibers pass modeling shows reciprocal effective communication between the right through the superior cerebellar peduncle (SCP) and the thalamus. From posterior STS and the left lateral cerebellar lobule crus I during visual [37], copyright © The Author 2012. Published by Oxford University processing of biological motion (BM) that modulates the back connection Press from the cerebellum to the STS. Adapted from [92], Copyright © 2011 for biological motion specificity in this series of studies) [91]. possibility for structural connection between the temporal In contrast, sensitivity was not impaired in patients with cortex and cerebellum had been detected by diffusion tensor lesions to the medial left cerebellum. In accord with lesional imaging (DTI) in non-human primates and humans [25]. By data, fMRI in a homogeneous group of healthy human adults using high-resolution acquisition sequences and optimized indicated activation of the left lateral cerebellar lobules crus I processing, our latest DTI work indicates a bidirectional struc- and VIIB [92]. Convergent lesion and brain imaging findings tural loop between regions in the left cerebellar lobule crus I provide reliable evidence in favor of involvement of the left and right STS that were functionally defined during visual lateral cerebellum in visual processing of human locomotion. processing of biological motion [37] (see Fig. 4c). Moreover, dynamic causal modeling demonstrated bidirec- In neuropsychiatric conditions such as schizophrenia or tional task-related effective connectivity between the left lat- autistic spectrum disorders (ASD), impaired biological motion eral cerebellar lobule crus I and the right STS during body processing [99, 113, 114] and altered cerebro-cerebellar con- motion perception [92](see Fig. 4b). The findings suggest that nectivity [115, 116] represent two major characteristics. Yet the cerebellum interacts with the cortical structure considered the relationship between these characteristics has not been as a hub of the neural network subserving visual processing of experimentally investigated. Reciprocal loops between the biological motion [105–107]. This may account for effects of cerebellum and STS in visual processing of body motion left lateral cerebellar lesions on visual tuning to biological may account for lower STS response to biological motion in motion [91]. children with ASD [117] and help to explain how social While closed cerebellar loops with the frontal and parietal deficits relate to disintegrity of the left superior cerebellar cortices are thought to underlie a variety of cognitive functions peduncle [118] hosting the back connection from the cerebel- [110], direct communication between the temporal cortex and lum to the STS [37]. Cerebellar involvement in biological cerebellum during a visual perceptual task had not been pre- motion processing instigates further research on social brain viously shown. Neuroanatomical evidence in non-human pri- networks in neuropsychiatric conditions. mates points to direct projections from the STS to the pons [8, In a nutshell, the left lateral cerebellum appears to be 9, 14, 111, 112] and from the pons to the cerebellum [27, 108]. strongly involved in visual processing of biological motion However, there has been lacking evidence for a back connec- [91, 92]. This engagement occurs likely through direct recip- tion from the cerebellum to the STS. Resting state fMRI rocal communication with the right STS [37, 92], a keystone analyses (see the section by Dr. Habas, “Resting-State Func- of brain networks for body motion processing and visual tional Connectivity Between Cerebellum and Sensory Sys- social cognition [99, 105–107]. Both specificity of deficits in tems (C. Habas)”) indicated possible functional connectivity patients with cerebellar lesions and network topography in between the cerebellum and temporal cortex [33–35]. A healthy adults suggest that cerebellar engagement in 206 Cerebellum (2015) 14:197–220 biological motion processing and action observation goes Recent electrophysiological analyses of the vestibulo- beyond a general role of the cerebellum in visual motion cerebellum and vestibular sensory pathway of monkeys have processing [86, 119; see also the section by Drs. Baumann provided important insights into the specific neural computa- and Mattingley, “The Role of the Cerebellum in Visual and tions underlying the integration of multimodal information Auditory Processing (O. Baumann and J.B. Mattingley)”]. required for self-motion perception. Recent data indicate a remarkable potential for recovery of First, to generate an accurate perception of our motion visual body motion processing following neurosurgical left relative to the world, the brain must continuously account cerebellar lesion removal and suggest that reorganization in for the omnipresent force of gravity. The brain constructs the cerebellum may trigger topographic shifts in the com- internal models of the world’s physical laws to dissociate tilt municating superior temporal areas [120]. The exact func- from translation by combining inputs from the vestibular tion of the cerebellum within the circuitry for perception of otoliths (which detect linear motion for both movements) with biological motion needs further clarification. Engagement inputs from semicircular canals (which detect rotational mo- of both the left cerebellum and right STS has been reported tion, and thus only respond to tilts) [132]. Consistent with this in emotion recognition through body motion [121], detec- proposal, single nodulus-uvula neurons create an internal tion of social interaction and animacy attribution in Heider- model that accounts for the physics of our world. Notably, and-Simmel movies depicting geometric shapes [122–124], neuronal responses to rotations are modulated as a function of imitation [125], and audiovisual integration ([93]; see sec- head orientation relative to gravity (reviewed in [127]) and tion of Drs. Baumann and Mattingley, “The Role of the different subclasses of Purkinje cells encode head translation Cerebellum in Visual and Auditory Processing (O. versus tilt [133]. This representation of translation could po- Baumann and J.B. Mattingley)”). Effective connectivity tentially be combined with the visual and proprioceptive input between the cerebellum and STS during animacy attribution to provide an estimate of heading direction that is based on has recently been demonstrated [124]. Further studies are information from multiple sensory systems. needed to clarify whether and how communication between Second, to perceive body motion independently of head the cerebellum and STS might underlie other social cogni- motion, the brain must compare vestibular and neck-related tive functions, and to address compensatory potential in inputs. Direct evidence for this computation has been revealed congenital, degenerative, and focal cerebellar affections. in the output of the cerebellum, at the level of the neurons in the most medial of the deep cerebellar nuclei (i.e., fastigial), which comprises two distinct populations of neurons. One neuronal population responds to both externally applied ves- The Cerebellum and Perception: The Role tibular and neck-proprioceptive stimulation, and encodes of the Cerebellum in Self-Motion Perception (K.E. Cullen) body-in-space motion. The other neuronal population only responds to externally applied vestibular inputs and encodes The cognitive representation of self-motion is vital to our head-in-space motion [134]. Notably, the convergence of everyday activities. For instance, walking down a busy city vestibular and proprioceptive inputs in body coding cerebellar street requires an accurate estimate of our own motion relative neurons is non-linear [134] and likely underlies the transfor- to objects in the surrounding complex, three-dimensional mation of vestibular signals from a head to a body reference environment. Self-motion requires the integration of sensory frame in the deep cerebellar nucleus [135, 136]. information from multiple systems including vestibular (head Finally, to ensure perceptual stability in everyday life, our motion), visual (optic flow), proprioceptive, and somatosen- brains must continually distinguish between self-motion that sory (body motion), as well as efference copy motor command is the result of our own (active) movements versus externally signals (reviewed in [126]). applied (passive) motion. Theoretically, the computation of There is strong evidence that the cerebellum, and, in par- passive motion requires a comparison between an internal ticular, the vestibulo-cerebellum, makes vital contributions to estimate of the sensory consequences of active self-motion self-motion perception. First, it has long been known that (i.e., forward model) and the actual sensory feedback lesions of the nodulus and uvula (Larsell’s lobules X and (reviewed in [126]). Cerebellar output neurons dynamically IX) alter the temporal and three-dimensional spatial process- encode this difference during self-motion; fastigial neurons ing of vestibular information (reviewed in [127]). More re- are insensitive to active motion and encode an explicit repre- cently, it has been further shown that visually induced illu- sentation of passively applied self-motion [137]. Specifically, sions of self-motion preferentially activate these same lobules the two distinct fastigial nucleus populations (described in the [128, 129] and that self-motion perception is diminished in paragraph above) selectively and dynamically encode passive patients with midline lesions impacting these regions [130, head and body motion relative to space. Moreover, our evi- 131]. Thus, the vestibulo-cerebellum is thought to be required dence to date suggests that this cerebellar-dependent mecha- for computing the internal representation of self-motion. nism uses an internal model of the expected sensory Cerebellum (2015) 14:197–220 207 The perception of pain itself is a complex subjective experi- ence that incorporates sensory, affective, and cognitive com- ponents. Though neuroimaging studies indicate that the cere- bellum responds to noxious stimuli, its functional relevance in relation to these different dimensions is only starting to gain attention. Ascending Nociceptive Input to the Cerebellum Well-controlled studies of pain often use acute experimental stimuli to activate nociceptive pathways, the physiological processes underlying pain perception. Nociceptors are prima- ry afferents that respond to high threshold mechanical and heat stimuli, and can also respond to chemical stimulation, such as during inflammation. Two major categories of noci- ceptive afferents have been classified: A-delta and C-fiber Fig. 5 The cerebellum integrates sensory input (green boxes) from nociceptors. A-delta nociceptors are thinly myelinated and multiple systems including: (1) the vestibular, (2) visual, (3) fast conducting (>2 m/s), while C-fiber nociceptors are unmy- proprioceptive and somatosensory, as well as from (4) motor efference copy signals. Cerebellar output neurons send ascending projections to the elinated and slower conducting (<2 m/s). Electrophysiological thalamus, hippocampus, and superior colliculus, which in turn connect studies in rodents and cats indicate that stimulation of cutane- the cerebellum to numerous cortical regions (red boxes) that mediate ous and visceral nociceptors, in the form of A-delta and/or C- spatial navigation and voluntary motor control fiber primary afferents, can activate and modulate Purkinje cell activity in the cerebellum [141, 142]. At least two possible nociceptive spinocerebellar pathways have been proposed: (1) consequences of active head motion to selectively cancel a spino-olivocerebellar pathway that conveys A-delta and C- responses to active motion. fiber nociceptive afferent input to Purkinje cells in the cere- In summary, computations in the vestibulo-cerebellum bellar anterior lobe ipsilateral to stimulation [143]and (2)a underlie the transformation of input signals into repre- spino-pontocerebellar pathway conveying C-fiber nociceptive sentations that are essential for self-motion perception input to Purkinje cells in the cerebellar vermis [144]. Details (Fig. 5). Interestingly, these same cerebellar-dependent regarding these putative pathways have been vastly computations likely also contribute to mapping spatial understudied. representation in the hippocampus (Fig. 5, ascending pathwayinred). Notably, ‘place cell’ tuning is impaired Descending Cortical and Subcortical Input to the Cerebellum in mutant mice with cerebellar function deficits [138]. The cerebellum likely shapes the directional tuning of In addition to afferent input, the cerebellum receives input place cells via indirect projections from the deep cere- from brain areas associated with nociceptive processing, in- bellar nuclei. Moreover, ascending projections terminate cluding cognition, affect, and motor function [141]. Our cur- in regions of the thalamus [139]knowntoterminate in rent understanding of the neural basis of pain and its modula- parietal cortex, a region that is vital for spatial naviga- tion includes the somatosensory cortices, periaqueductal gray, tion,aswellasmotor andpremotor cortex[140]. Future anterior cingulate cortex, dorsolateral prefrontal cortices, basal work in monkeys and mice using both passive and ganglia, hippocampus, hypothalamus, and the amygdala active motion are needed to fully understand the impact [145], all of which have connectivity with the cerebellum [1, of the cerebellum on how the hippocampus and cortex 96]. With the cerebellum receiving both descending informa- shape spatial navigation. tion from other brain areas and ascending nociceptive infor- mation from the spinal cord, the structure is ideally positioned to influence, or to be influenced by, the processing of pain. Pain and the Cerebellum (R.J. Borra and E.A. Moulton) Neuroimaging Responses to Pain in the Cerebellum The cerebellum is one of the most consistently responsive brain structures to painful stimuli [141]. While our classical A meta-analysis of 47 neuroimaging studies featuring exper- understanding of this structure suggests that it is involved in imental pain revealed specifically localized responses within the motor response to pain, contemporary thinking indicates the cerebellar vermis and bilaterally in the posterior hemi- spheres [141]. The spatial extent of vermal activation spanned that it may have a more direct role in the processing of pain. 208 Cerebellum (2015) 14:197–220 across vermal lobules III, IV, and V, while the bilateral hemi- studying cerebellar-dependent timing [150, 151]. This form spheric activation spanned from hemispheric lobule VI to crus of learning is only adaptive if the animal is able to represent I. Using the same method of meta-analysis, a similar pattern of the temporal relationship between two sensory events, the activation was observed across 16 neuroimaging studies fea- conditioned and unconditioned stimuli. Importantly, the con- turing pathological pain, in the form of spontaneous pain or ditioned response persists following lesions of the cerebellar aggravation of a clinical condition. cortex but loses its adaptive timing [152](see Fig. 6a). Sen- Though pain neuroimaging studies are not typically de- sory timing as a constraint on motor control is also evident in signed to evaluate the physiological significance of cerebellar many tasks involving volitional movements. To intercept a responses, a few notable studies have focused in on this moving object, the movement has to anticipate the trajectory structure in the context of pain. Helmchen and colleagues of the object. Patients with cerebellar lesions have great diffi- used fMRI to find that activation in hemispheric lobule VI culty with such tasks [153]. Mice lacking genes associated with and in the anterior vermis varied with subject reports of pain cerebellar-dependent plasticity are selectively impairedinan intensity, though only when the stimuli were self-administered operant task that requires using precise sensory timing to restrict [146]. The authors suggested that these cerebellar regions movement latencies [154]. could reflect pain perception and are involved in signaling The preceding examples highlight a critical cerebellar role the expected sensory consequences of pain. In another study, in using sensory information to time movement. The reverse fMRI of trigeminal neuropathic pain elicited by brushing and situation, where movement is used to anticipate and modulate heat showed responses in crus I, crus II, and lobule VIIB that sensory information, is also cerebellar dependent, at least were not evoked by the non-painful control stimuli [147]. when the events are of a limited temporal extent. We have Recent neuroimaging evidence suggests that certain cere- general consensus that the cerebellum uses a forward model to bellar responses during pain may reflect multi-modal aversive generate a prediction of the expected sensory consequences of processing. An fMRI study found that noxious heat and the an action [155]. Kotz and colleagues [156]provide aparticu- passive viewing of unpleasant pictures activated overlapping larly compelling EEG example. The early N100 response regions of the cerebellum: hemispheric lobules VI, VIIB, and evoked by an auditory stimulus is markedly attenuated when crus I [148]. Further analysis revealed that these areas of the tones are triggered by a volitional action compared with functional overlap were significantly inversely correlated with when the tones are externally triggered. This attenuation is activation in the anterior hypothalamus, subgenual anterior essentially absent in patients with focal cerebellar lesions of cingulate cortex, and the parahippocampal gyrus. These find- the left or right hemisphere, with the sensory response similar ings suggest that responses in these cerebellar regions are not for self-triggered and externally triggered actions (see specific to pain processing but appear to apply to other aver- Fig. 6b). Forward models, as a form of prediction, have been sive sensory and affective experiences as well [149]. Howev- er, other functions related to pain aside from aversion may also employed to describe brain function more generally [157]. A be processed in the cerebellum, as areas that responded to challenge is to specify the conditions that distinguish noxious heat and not to aversive pictures were also identified cerebellar-dependent and cerebellar-independent forward including crus II. Further study is required to determine the models. One possibility is that, as with classical conditioning, functional topography of the cerebellum as it relates to pain the cerebellar domain is defined by temporal constraints, and its different sensory, affective, and cognitive components. situations in which the predictions require some form of precise temporal representation. In one oft-cited example, the tickling sensation from self-generated movements be- comes more intense when delays are introduced between the Sensory Processing and the Cerebellum: Timing action and the somatosensory stimulation [158]. Similarly, (R.B. Ivry) learning rates are dramatically reduced with delayed feedback during visuomotor adaptation [159]. Movement dynamically incorporates sensory information and The strongest evidence for a critical role of the cerebellum anticipates the sensory consequences of the action (see also in sensory timing comes from tasks that do not entail overt the section by Dr. Bower, “Is the Cerebellum Sensory for movement (see also the section by Drs. Baumann and Motor’s Sake, or Motor for Sensory’s Sake? (J.M. Bower)”). Mattingley, “The Role of the Cerebellum in Visual and Audi- While this is a general feature of motor control, there is tory Processing (O. Baumann and J.B. Mattingley)”). Re- consensus of a cerebellar dependency on tasks that impose search here falls into three general domains. First are tasks precise temporal constraints. A prominent feature of cerebellar examining how the cerebellum responds to temporal regular- ataxia is the loss of the fine temporal patterning that is char- ities, or perhaps more telling, violations of temporal expec- acteristic of skilled movement. Experimentally, eyeblink con- tancies. Tesche [160] compared evoked MEG responses to ditioning has proven to be an exquisite model system for periodic (predictable) tactile stimuli or epochs in which the Cerebellum (2015) 14:197–220 209 Fig. 6 Cerebellum and sensory timing. a Adaptive timing of conditioned eye blink response is abolished following infusion of picrotoxin, an agent that disrupts input from cerebellar cortex to deep cerebellar nuclei. Courtesy of Michael Mauk. b Patients with focal cerebellar lesions fail to show attenuated ERP response to self-generated sounds compared with externally produced sounds. Adapted from [156]. c Patients with cerebellar degeneration (SCA6) exhibit selective deficit on time perception tasks that require interval timing (Var, Fix) while spared performance on tasks that require beat-based timing (Reg, Iso, Met). Adapted from [165]. d Cerebellar grey matter volume is correlated with perceptual acuity on time discrimination task, relative to a color discrimination task. Adapted from [173] stimulus was withheld (prediction violations). Whereas the cerebellar BOLD response was larger in the latter compared evoked response in somatosensory cortex was stimulus- with the former. Converging evidence comes from a study locked and independent of predictability, the cerebellar re- showing that patients with cerebellar pathology are impaired sponse was anticipatory, leading the expected onset of the in adapting to velocity perturbations in this task [58]. stimulus. Moreover, it was markedly larger following a viola- Third, and perhaps most direct, are studies of duration tion, consistent with the idea that the cerebellum was sensitive discrimination. Ivry and Keele [88] provided the first evidence to temporal prediction violations. Further support for this idea of a “pure” sensory timing deficit in patients with cerebellar comes from fMRI work showing larger cerebellar activation pathology. The patients were impaired in judging the duration to visual stimuli with unpredictable timing (e.g., [161]) as well of an auditory stimulus but showed normal performance in as a study in which an early ERP signal to deviant auditory judging stimulus loudness. This finding has been confirmed in stimuli was found to be abnormal in patients with cerebellar various studies over the past 25 years, including one study in degeneration [162]. which testing was restricted to a large group of patients with The second domain involves studies of velocity perception. SCA6, a condition in which the pathology is relatively re- Cellular activity in the posterior cerebellum is sensitive to stricted to the cerebellar cortex [165](see Fig. 6c), and studies stimulus motion (see sections by Dr. Cullen on “The Cerebel- with healthy individuals in which cerebellar function has been lum and Perception: The Role of the Cerebellum in Self- transiently disrupted by TMS [166, 167]. There is general Motion Perception (K.E. Cullen)”, and Drs. Sokolov and consensus that cerebellar contributions to sensory (and motor) Pavlova, “Cerebellar Involvement in Biological Motion Pro- timing are most pronounced with relatively short intervals cessing (A.A. Sokolov and M.A. Pavlova)”). It is possible that (less than 1 s) and in the representation of intervals (either these signals are related to preparation of potential eye or body absolute or relative as in state estimation models) rather than movements. However, a causal contribution to perception more complex temporal relationships (e.g., rhythm). The few comes from psychophysical studies showing that patients with negative results on duration discrimination are also informa- cerebellar pathology are impaired in visual motion discrimi- tive: They have involved patients with unilateral lesions [168, nation [86, 163]. Moreover, the cerebellar contribution appears 169], suggesting that a single intact cerebellar hemisphere to be most critical when the motion perception task requires may be sufficient to support sensory timing [170]. The func- time-based judgments. O’Reilly [164] used a task in which a tional neuroimaging literature on duration perception has moving stimulus disappeared behind an occluder. When the proven more difficult to decipher [171], especially since many stimulus reappeared, the participant had to judge if there had studies do not provide adequate coverage of the cerebellum. Interestingly, three recent structural MRI studies report a been a deviation in direction (spatial) or speed (temporal). The 210 Cerebellum (2015) 14:197–220 positive correlation between measures of cerebellum volume Mismatch negativity (MMN) studies in subjects with cer- and temporal acuity in healthy individuals [172–174] ebellar damage in the somatosensory [182] or auditory [162] (see Fig. 6d). domain have confirmed this hypothesis. MMN is believed to This is not to say there is consensus for the uniqueness be generated by an automatic cortical change-detection pro- of the cerebellum in sensory timing. Indeed, there is con- cess that is activated by differences between current and prior sensus that the cerebellum is not the sole structure capable inputs. When the MMN protocol is applied to subjects with of representing temporal information. The challenge re- cerebellar lesions, the MMN response is absent or abnormal. mains to develop more analytic tasks and models that Per the long-standing model in which the cerebellum acts as a provide better specification of the various operations re- comparator [183], it has been proposed that, in the cerebellum, quired in tasks that require precise temporal processing. actual input and preceding stimuli are compared, and discor- Nonetheless, the cerebellar timing hypothesis [98]has dances are identified. If the incoming stimulus corresponds to proven to be of considerable utility for exploring the the predicted stimulus, cerebellar output is minimal; if a function, structure, and physiological of the cerebellum discrepancy–error signal is detected, the activity in the cere- in motor control and beyond. bellum increases and a large area of the cerebral cortex is alerted by enhancing its excitability (Fig. 7). We developed a “sequence detection model” to describe the operational mode of cerebellar processing not only in somato- The Cerebellum in Predicting Perceptual Events sensory [182], but also in visuospatial [184] and cognitive (M. Leggio and M. Molinari) domains [185]. Cerebellar patients were impaired specifically in the recognition of spatial sequences when tested on a visuo- Perception can be considered the result of interactions in time spatial serial reaction time task [184]. Results of visuospatial between a dynamic mind and a dynamic world. To achieve tests demonstrated that subjects with cerebellar damage were mind-world synchronization, our perceptual systems must impaired specifically with regard to sequence recognition, even constantly tune themselves to an ever-changing environment. to a greater extent than sequence execution [184]. Furthermore, Perceptual tuning, like the sensorimotor tuning that is needed by forcing the declarative knowledge of the spatial order, it was for smooth movement control, can be obtained only if predic- possible to improve performance significantly. Similar findings tion capabilities are embedded in the process [175]. Moreover, have been reported by several groups [186–191], supporting predictive processing represents a fundamental principle of cerebellar function in extracting sequential order information neural computations in the brain [176]. from incoming sensory information [184]. Many groups have attempted to identify the neural bases of Subjects with cerebellar damage also develop impairments in foresight, and despite considerable ongoing debate, a consen- cognitive sequencing [192]. We analyzed prediction ability in sus exists on the importance of the cerebellum in prediction patients with cerebellar damage who performed a cognitive task [177]. To make the matter even more interesting for cerebellar in which predictability was based primarily on abstract/spatial, scientists, data are accumulating on the significance of the behavioral/visual, or behavioral/linguistic sequence information cerebellum for sensory processing and in optimizing percep- [192]; in this task, sets of cartoon-like drawings that reproduced tion [58]. Perceptual optimization and prediction of incoming behavioral sequences were to be placed in the correct order. The sensory information have been suggested to be effected by patients were impaired in sequencing events in all domains, sequence processing in cerebellar circuits [58, 178, 179]. developing domain-related specificity, based on the side of the Using magnetoencephalographic recordings, Tesche and cerebellar lesion. Thus, although no specific sequencing locali- Karhu [160] demonstrated that cerebellar activity is enhanced zation can be identified, sequencing processing can be found in after an unpredictable omission is inserted into a regular train the different cerebellar functional domains. This impairment of somatosensory stimuli. As a result, no activity is present in suggests difficulties in perceiving the depicted behavior correctly. the parietal cortex, whereas a notable response develops in the This evidence is consistent with difficulties that are encountered cerebellum. Consequently, it can be argued that the cerebel- in tuning behavior and the environment correctly not only after lum detects the absence of a somatosensory stimulus to a cerebellar damage [19], but also in behavioral pathologies, such greater extent than its presence. This response to the absence as autism and schizophrenia, disorders that have been linked to of a stimulus can be understood only as an indication that cerebellar abnormality [193, 194]. something that is expected does not appear [180]. If a sensory The hypothesis that pattern detection and prediction repre- pattern is recognized, it is possible to predict the sequence of sent a specific role in cerebellar function in perception is events and consequently anticipate each one [181]. Thus, in appealing, and compelling data from various sources support predicting incoming sensory information, the cerebellum gov- the sequence detection model of impaired cerebellar percep- erns the detection of the absence of an expected stimulus and tion. Furthermore, the perceptual deficits that are observed in the appearance of an unexpected stimulus. schizophrenia [195, 196] and autism [197, 198] resemble Cerebellum (2015) 14:197–220 211 Fig. 7 Sequence detection model of prediction. If sensory events appear in a fixed sequence repeatedly in a short time, the sensory sequence is implicitly memorized a which allows cerebellar circuits to compute a prediction for forthcoming perceptual events b.If the prediction holds c, a signal is sent to the cerebral cortex to alert selective brain areas, which become activated prior to the realized event and are thus better suited to process the incoming stimulus. If the prediction fails d, an alertsignalis sent, andbrain activation is more widespread, accelerating the processing of salient sensory information by the changing events and attuning the behavioral response to the new environment cerebellar dysfunctions. Notably, cerebellar pathogenic mech- behavioral including human studies. First in a series of imag- anisms have been hypothesized to mediate schizophrenia ing experiments, we demonstrated, as the hypothesis predicts, [199] and autism [200], and the existence of cerebellar-like that activity in the human cerebellum [51, 206] and related sequence detection deficits [201, 202] is additional support for structures [207] is substantially greater when fingers are used the cerebellar pathogenic theories of these diseases. in a tactile discrimination task. A meta-analysis of neuroim- aging data then generalized this result to the auditory system, suggesting larger and more spatially extensive activations Is the Cerebellum Sensory for Motor’s Sake, or Motor during discriminative auditory tasks [208], a result subse- for Sensory’s Sake? (J.M. Bower) quently confirmed using PET [90]. Importantly, the PETstudy also supported a further important prediction of the sensory acquisition hypothesis, namely that cerebellar activity should This title is the same as a paper published more than 15 years ago describing our hypothesis that the cerebellum controls the increase with task difficulty, i.e., when better control of the quality of sensory data is likely more important [97]. A similar acquisition of sensory data [203], an idea first proposed even 10 years earlier in a paper exploring the spatial structure of the result has been reported independently in a combined human visual and auditory imaging study [89]. extensive peri-oral tactile representations in the cerebellum of rats: While human imaging data can be suggestive of brain “… we suggest that tactile regions of the cerebellum are function, an important test of any functional hypothesis is its involved in controlling the movements specifically associated ability to predict behavioral results. In this case, it has long with active tactile exploration …to coordinate the use of been a central tenet of cerebellar descriptions that the structure sensory structures so that the highest quality sensory informa- has no influence on sensory perception [209]. However, be- tion is being obtained by the rest of the nervous system during cause sensory perception is based on the quality of sensory data, we predicted that impairment of cerebellar function the exploratory process. By monitoring the acquisition of sensory information and adjusting motor performance accord- should have sensory perceptive consequences [203]. Consis- tent with this prediction, we have shown that humans with ingly, cerebellar circuits would be expected to substantially improve the efficiency of sensory processing by the rest of the cerebellar degenerative disease have significantly poorer thresholds for pitch discrimination [57]. Other studies in au- nervous system.” (p. 776, [204]) dition [86], somatosensation [210, 211], proprioception, and vision [212] have now also demonstrated cerebellar-related Evidence in Support primary sensory deficits, which have also been reported using higher order tasks like speech [213], motion detection [214], While a model-based re-analysis of cerebellar cortical net- works alsosupportsthishypothesis[205], this review will analysis of temporal sequences [178], as well as the general perception of time [167, 215]. focus on supporting experimental results from more 212 Cerebellum (2015) 14:197–220 Finally, while human psychophysical and imaging studies, but instead because sensory information is temporally coded properly designed, can test functional hypotheses, linking at the neuronal level [228], and therefore experimental these hypotheses to actual physical computational mecha- manipulations of expected timing relationships in presented nisms still requires the use of animal models [205]. While stimuli are likely to evoke stronger cerebellar effects. the large majority of animal studies exploring the functional 3. The cerebellum is invoked in proportion to the need for significance of the extensive sensory projections to the cere- sensory vigilance. Another important prediction of our bellum continue to frame the results in the context of tradi- hypothesis is that cerebellar involvement will scale as better tional motor control theories [216], a recent behavioral study controlled sensory data are required [90], making it impor- in rats has demonstrated that optogenetic stimulation of the tant to evaluate task difficulty when considering cerebellar- cerebellum specifically disrupts the use of the whiskers during related sensory effects [58, 229, 230]. Interestingly, numer- active touch [217]. These authors specifically conclude that ous cerebellar studies already employ masking sensory noise their results support a role of the cerebellum in the “optimiza- to evoke larger cerebellar responses [89, 92]orreveal be- tion of sensory data acquisition” (p. 6, [217]). havioral deficits [163]. Overcoming the consequences of sensory noise either applied externally or self-generated Implication for Theories of Cerebellar Sensory Function [231] we predict will especially increase requirements for cerebellar control. This effect also confounds the interpreta- With growing evidence that the cerebellum plays some role in tion of sensory stimuli like pain [141], whichontheir own sensory function, it is time to fully reconsider cerebellar increase subject vigilance [148, 232], as well as studies of function from a sensory point of view: mechanisms like attention [233–235]. 4. The cerebellum is a support structure. Perhaps the most 1. Re-interpreting cerebellar involvement in motor control. important implication of the sensory hypothesis is that the It has been known for more than 150 years that lesions of cerebellum performs a more internal than external func- the cerebellum disrupt movement [218, 219] with the tion. Instead of itself contributing directly to sensory majority of cerebellar theories accordingly focused on perception, the influence of the cerebellum is predicted mechanisms of direct motor control [220]. In contrast, to be indirect, facilitating the computational efficiency of the sensory data acquisition hypothesis proposes that the rest of the brain, including cerebral cortex [163]. To cerebellar effects on movement are an indirect conse- quote again from 25 years ago: quence of disrupting the sensory data on which motor behavior depends [97]. This prediction is consistent with “It has been largely accepted that the flocculus of the recent evidence that cerebellar patients have difficulty cerebellum is involved in adjusting the gain of the discriminating proprioceptive stimuli [210] and that a vestibulo-occular reflex to assure a minimal slip of significant component of cerebellar ataxia results from images on the retina during head movement [236, the inability of patients to perceive environmental insta- 237]. Psychophysical experiments demonstrate that bilities [221]. For this reason, it is critically important that more than 3°/sec of retinal slip starts to significantly motor-related studies, perhaps especially those involving degrade visual acuity and thus the ability of the visual purported motor learning [222], control for cerebellar system to process sensory information [238]. Thus the effects on primary sensory data. proposed role of the cerebellum, in VOR control, is to 2. Removing the legacy of cerebellar motor control theories. assure that the highest possible quality of visual infor- At present, most explanations for cerebellar involvement mation is provided to the visual system. In principle, this in non-motor-related behaviors assume that evolution has role is analogous to the role we are suggesting for lateral adopted cerebellar motor control computational mecha- tactile regions of the cerebellar cortex.” (p. 776, [204]). nisms to non-motor tasks [58, 73, 92, 211, 220, 223–226], including, for example, a presumed general role for the For sensory systems like vision, audition, olfaction, and cerebellum in timing not only of muscle activations dur- somatosensation, which in humans involve the largest part of ing movement but also of sensory perception [98, 227]. the cerebellum, the ‘support system’ status of cerebellum also While our analysis of cerebellar cortical circuitry ques- suggests a different interpretation of the important relationship between the cerebellum and the cerebral cortex. While in the tions the circuitry-based evidence for the original timing hypothesis [205], we do expect that any disruption in traditional motor-control context, the influence of the cerebral cortex on the cerebellum is generally described as sensory data acquisition control may very well be partic- ularly apparent with tasks involving precise timing (see implementing a kind of forward model to (quoting Dr Ivry in this article) “generate a prediction of the expected sensory the section by Dr. Ivry, “Sensory Processing and the Cerebellum: Timing (R.B. Ivry)”). This is not, however, consequences of an action” (see also the section by Dr. Paulin, because the cerebellum itself implements a timing function, “Evolutionary Perspectives on Cerebellar Function (M.G. Cerebellum (2015) 14:197–220 213 Paulin)”), in our view, the influence of the cerebral cortex on Summary and Conclusions the cerebellum provides contextual information related to the expected use of the sensory data by cerebral cortex. We don’t The aim of this consensus paper is to capture the range of think that such a function explicitly involves a ‘prediction’ as experimental approaches and theoretical models that have much as it does a continuous stream of contextual informa- contributed to our current understanding of the influence of tion. In fact, although again beyond the scope of the current the cerebellum on perceptual processes. Contributions from commentary, our analysis of cerebellar cortex suggests that its fourteen experts, spanning a range of methodological ap- circuitry specifically places information arising from particu- proaches and with different theoretical views, have been lar sensory receptors (e.g., the upper lip) in the context of other brought together to provide an up-to-date snapshot of thinking sensory surfaces involved at the same time in sensory data on this topic. acquisition (e.g., the lower lip). We have proposed that cere- The outcome of this project indicates that no single, coher- bellar output (through direct projections to the midbrain and ent model has yet emerged regarding the mechanisms by brain stem motor centers as well as potentially through motor which the cerebellum may influence perception. Nonetheless, regions of cerebral cortex) then makes subtle relative adjust- it is important to assemble the empirical data, showing the ments in the position of tactile sensory surfaces to optimize the association of the cerebellum with a wide range of perceptual information content. A recent analysis of the influence of the systems including those related to vision, audition, touch, cerebellum on whisking in rats supports this prediction [217]. proprioception, self-motion perception, and nociception. The Similarly, we have proposed that the cerebellum also likely possible anatomical and physiological underpinnings of this modulates the cochlear outer hair cells during auditory data broad influence was reviewed by Dr. Schmahmann, acquisition. In fact, we have suggested that the cerebellum documenting significant cerebellar connection with sensory, plays the same role for all sensory systems. as well as associative and paralimbic, areas of the cerebrum. 5. Implications for human disease. Finally, the most exciting These findings are corroborated by human neuroimaging application of this sensory focused hypothesis may be to studies, which show that fMRI resting-state signals in the human health and disease. Although understudied, it has cerebellum correlate significantly with those in visual and been known for more than 150 years that motor control auditory cortices in the cerebrum (see the section by Dr. can recover after cerebellar cortical lesions [239, 240]an Habas, “Resting-State Functional Connectivity Between Cer- effect also now demonstrated for presumed ‘cognitive’ ebellum and Sensory Systems (C. Habas)”). Second, a number function [241–243]. The sensory hypothesis attributed of the commentators described clinical studies that show how this recovery to the eventual adaptation of the rest of the cerebellar lesions can lead to deficits in a diverse set of brain to less well-controlled sensory data [203]. Evidence perceptual tasks, including visual motion perception, auditory has also been growing that the cerebellum plays a role in pitch perception, self-motion perception, biological motion autism spectrum disorders (ASD), although there is no perception of others, time perception, and the recognition of consensus for the mechanism [244]. In the context of our perceptual sequences (see sections by Drs. Baumann and hypothesis, the relationship is quite direct, with ASD seen Mattingley, “The Role of the Cerebellum in Visual and Audi- as a behavioral adaptation to a general and overwhelming tory Processing (O. Baumann and J.B. Mattingley)”; Drs. lack of control over the process of sensory data acquisi- Pavlova and Sokolov, “Cerebellar Involvement in Biological tion. From this perspective, therapies that focus on repet- Motion Processing (A.A. Sokolov and M.A. Pavlova)”;Dr. itive behaviors in highly controlled sensory environments Cullen, “The Cerebellum and Perception: The Role of the with specific emphasis on sensory integration [245] Cerebellum in Self-Motion Perception (K.E. Cullen)”;Dr. would, we suggest, establish sensory conditions making Ivry, “Sensory Processing and the Cerebellum: Timing (R.B. it easier for the brain to learn to compensate for the lack of Ivry)”; Drs. Leggio and Molinari, “The Cerebellum in stable sensory data. It may even be worth considering Predicting Perceptual Events (M. Leggio and M. Molinari)”; whether the apparent increasing incidence of ASD could and Dr. Bower, “Is the Cerebellum Sensory for Motor’s Sake, be attributable to sensory over-stimulation of children or Motor for Sensory’s Sake? (J.M. Bower)”). Third, human before the late developing cerebellum is fully functional. neuroimaging studies have consistently shown reliable cere- In summary, there is no question that the evidence is bellar activation during performance of a range of perceptual growing for some kind of cerebellar involvement in tasks, independent of any motor-related activity of observers mechanisms of sensory function. However, instead of (see sections by Drs. Baumann and Mattingley, “The Role of assuming a direct role in these mechanisms borrowing the Cerebellum in Visual and Auditory Processing (O. traditional cerebellar theories designed to explain motor Baumann and J.B. Mattingley)”; Drs. Pavlova and Sokolov, control, in our view, this new evidence should instead call “Cerebellar Involvement in Biological Motion Processing into question the historical view of the cerebellum as (A.A. Sokolov and M.A. Pavlova)”; Drs. Borra and Moulton, primarily a motor control device. “Pain and the Cerebellum (R.J. Borra and E.A. Moulton)”; 214 Cerebellum (2015) 14:197–220 and Dr. Bower, “Is the Cerebellum Sensory for Motor’s Sake, these hypotheses. Most of the current evidence is deliv- or Motor for Sensory’s Sake? (J.M. Bower)”). ered by human lesion and neuroimaging studies, methods In summary, it seems the answer to the question of whether that have provided valuable insights from a systems-level the cerebellum plays a role in perception is unequivocally perspective, but are of limited value in constraining affirmative. What remains to be determined is precisely how models at the level of microcircuitry. It is therefore essen- the cerebellum contributes to perceptual processes. tial to also explore the cerebellum’s involvement in per- Dr. Schmahmann sets the stage for functional hypotheses. ceptual tasks at the level of single neurons. Dr. Cullen’s Inspired by the cerebellum’s uniform neuroanatomical struc- research on the role of the cerebellum in self-motion ture and dense heterogeneous connectivity, he argues that we perception provides a compelling example. By recording should assume a constant computation—the universal cere- from individual cerebellar neurons, her research has shown bellar transform—that is applied to multiple domains of neu- that the cerebellum computes sensory prediction error sig- rological function determined by cerebellar connections. The nals that effectively distinguish between the sensory con- idea of a uniform computation is repeated in many of the other sequences of self-generated and externally produced ac- commentaries, although the specific form of the computation tions. These findings seem inconsistent with the conven- shows considerable variation. Building on comparative data tional view that the role of the cerebellum is restricted to from across the animal kingdom, Dr. Paulin suggests that the motor learning. cerebellum provides the ability to predict state trajectories of Finally, an important application of new knowledge arising dynamical systems. The ability to predict state trajectories of from research into the role of the cerebellum in perception is in the body and external targets is essential for agile motor the domain of human health and disease. The historical asso- control and can explain the obvious, classical symptoms of ciation of the cerebellum with “motor function” has limited cerebellar dysfunction. But state estimation can also provide appropriate consideration of its potential role in perceptual core capability for a variety of signal processing, decision- functions, in both health and disease. It is now apparent that making and control tasks, and this could explain newer evi- cerebellar lesions can lead to a range of behavioral, cognitive, dence about the cerebellum’s role in non-motor tasks. The affective, and perceptual impairments. In addition, psychiatric latest neuroimaging evidence for direct interaction between conditions that are characterized by perceptual and cognitive the cerebellum and temporal areas involved in visual motion (as well as motor) disturbances, including autism, schizophre- processing and body motion processing (MT/MST and STS), nia, and attention deficit hyperactivity disorder, are associated as presented by Drs. Baumann, Mattingley, Pavlova and with cerebellar pathology. The possibility of a cerebellar role Sokolov, appears to lend further support to this hypothesis. in the manifestations or pathogenesis of these conditions Similarly, Dr. Ivry’s hypothesis proposes a contribution of the is intriguing. Further research into the role of the cere- bellum in perceptual functions may help to advance our cerebellum to the analysis and prediction of sensory event timing in the sub-second range. Drs. Leggio and Molinari’s understanding of the mechanisms underlying these dis- hypothesis of the cerebellum’s role in perception shares the orders. Moreover, patients with isolated cerebellar in- central assumption that the cerebellum is involved in the sults, cerebellar tumors, and hereditary cerebellar degen- analysis and prediction of dynamic perceptual events. While erative disease will also benefit from a better under- Dr. Ivry focused here on a narrower view of prediction, events standing of the role of the cerebellum in perception. requiring precise timing in the sub-second range, Drs. Leggio To date, diagnostic evaluation and therapeutic interven- and Molinari take a broader view of prediction with their tions in patients with cerebellar disease have been lim- hypothesis that the cerebellum supports perception by ited to the striking deficits in the coordination of vol- extracting sequential order information from incoming senso- untary movements. Recognition of a cerebellar role in ry information. Clinical and neuroimaging studies not sensory processes helps to identify and treat potential only implicate the cerebellum in the analysis of dynam- perceptual deficits that may at present go unnoticed and ic stimuli, but also in less dynamic perceptual tasks untreated. In addition, further research on the compen- such as pitch discrimination and nociception. Dr. Bower satory potential of not only motor, but also perceptual urges us to consider that the cerebellar contribution cerebro-cerebellar networks after cerebellar damage may arises at an even earlier stage of processing, arguing advance both clinical management and understanding of that the cerebellum influences perception by controlling the cerebellar contribution to perception. the acquisition of sensory data, an idea that might This review is the first attempt to capture the variety of explain why cerebellar activity often increases with the current experimental approaches and theoretical models on difficulty of a perceptual task. the cerebellum’s role or influence on perception. By drawing While some of the described theories could be seen as together the diverse perspectives, we intend to stimulate sci- complementary, the challenge remains to develop more entific debate and increase interest in the cerebellum and its explicit experimental tests that can distinguish between complex functions. Cerebellum (2015) 14:197–220 215 Acknowledgments (1) Dr. Schmahmann’s work was supported in part 14. Schmahmann JD, Pandya DN. Projections to the basis pontis from by the MINDlink and Birmingham Foundations. (2) Dr. Baumann was the superior temporal sulcus and superior temporal region in the supported by an Australian Research Council (ARC) Discovery Early rhesus monkey. J Comp Neurol. 1991;308:224–48. Career Researcher Award (DE120100535) and Dr. Mattingley by an 15. Ungerleider LG, Desimone R, Galkin TW, Mishkin M. Subcortical ARC Australian Laureate Fellowship (FL110100103), the ARC-SRI projections of area MT in the macaque. J Comp Neurol. 1984;223: Science of Learning Research Centre (SR120300015), and the ARC 368–86. Centre of Excellence for Integrative Brain Function (ARC Centre Grant 16. Schmahmann JD, Pandya DN. Prelunate, occipitotemporal, and CE140100007). (3) Dr. Pavlova was supported by Else Kröner Fresenius parahippocampal projections to the basis pontis in rhesus monkey. Foundation (Grant P2013_127), the Reinhold-Beitlich Foundation, the J Comp Neurol. 1993;337:94–112. Berthold Leibinger Foundation, and the Heidehof Foundation (Grant 17. Fiez JA, Raichle ME. Linguistic processing. In: Schmahmann JD, 59073.01.1/3.13). (4) Dr. Borra was supported by funding from the Sigrid editor. The cerebellum and cognition. International review of neu- Juselius Foundation, the Instrumentarium Research Foundation, the Finn- robiology, vol. 41. San Diego: Academic; 1997. p. 233–54. ish Medical Foundation, the Paulo Foundation and the Academy of 18. Stoodley CJ, Schmahmann JD. Functional topography in the human Finland (270352). Dr. Moulton was supported by the National Institute cerebellum: a meta-analysis of neuroimaging studies. Neuroimage. of Health, USA (NIH/NCI R21CA185870). (5) Dr. Ivry was supported 2009;44:489–501. by the National Institute of Health, USA (NS084948 and NS074917). 19. Schmahmann JD, Sherman JC. The cerebellar cognitive affective syndrome. Brain. 1998;121:561–79. 20. Mariën P, Ackermann H, Adamaszek M, et al. Consensus paper: Conflicts of Interest The authors declare that no financial or personal language and the cerebellum: an ongoing enigma. Cerebellum. competing interests exist. 2014;13:386–410. 21. Picard N, Strick PL. Motor areas of the medial wall: a review of their Open Access This article is distributed under the terms of the Creative location and functional activation. Cereb Cortex. 1996;6:342–53. Commons Attribution License which permits any use, distribution, and 22. Vilensky JA, Van Hoesen GW. Corticopontine projections from the reproduction in any medium, provided the original author(s) and the cingulate cortex in the rhesus monkey. Brain Res. 1981;205:391–5. source are credited. 23. Aas J-E, Brodal P. Demonstration of topographically organized projections from the hypothalamus to the pontine nuclei: an exper- imental study in the cat. J Comp Neurol. 1988;268:313–28. 24. Haines DE, Dietrichs E. An HRP study of hypothalamo-cerebellar References and cerebello-hypothalamic connections in squirrel monkey (Saimiri sciureus). J Comp Neurol. 1984;229:559–75. 25. Ramnani N, Behrens TE, Johansen-Berg H, et al. The evolution of 1. Schmahmann JD. From movement to thought: anatomic substrates prefrontal inputs to the cortico-pontine system: diffusion imaging of the cerebellar contribution to cognitive processing. Hum Brain evidence from Macaque monkeys and humans. Cereb Cortex. Mapp. 1996;4:174–98. 2006;16:811–8. 2. Schmahmann JD. The cerebrocerebellar system: anatomic sub- 26. Allen GI, Tsukahara N. Cerebrocerebellar communication systems. strates of the cerebellar contribution to cognition and emotion. Int Physiol Rev. 1974;54:957–1008. Rev Psychiatry. 2001;13:247–60. 27. Brodal P. The pontocerebellar projection in the rhesus monkey: an 3. Schmahmann JD. The role of the cerebellum in cognition and experimental study with retrograde axonal transport of horseradish emotion: personal reflections since 1982 on the dysmetria of peroxidase. Neuroscience. 1979;4:193–208. thought hypothesis, and its historical evolution from theory to 28. Kelly RM, Strick PL. Cerebellar loops with motor cortex and therapy. Neuropsychol Rev. 2010;20:236–60. prefrontal cortex of a nonhuman primate. J Neurosci. 2003;23: 4. Schmahmann JD, Pandya DN. The cerebrocerebellar system. In: 8432–44. Schmahmann JD, editor. The cerebellum and cognition. San Diego: 29. Middleton FA, Strick PL. Cerebellar output channels. In: Schmahmann Academic; 2007. p. 31–60. JD, editor. The cerebellum and cognition. International review of 5. Stein JR, Glickstein M. Role of the cerebellum in visual guidance of neurobiology, vol. 41. San Diego: Academic; 1997. p. 61–82. movement. Physiol Rev. 1992;72:967–1017. 30. Buckner RL, Krienen FM, Castellanos A, Diaz JC, Yeo BT. The 6. Oscarsson O. Functional organization of the spino- and organization of the human cerebellum estimated by intrinsic func- cuneocerebellar tracts. Physiol Rev. 1965;45:495–522. tional connectivity. J Neurophysiol. 2011;106:2322–45. 7. Barmack NH. Central vestibular system: vestibular nuclei and pos- 31. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connec- terior cerebellum. Brain Res Bull. 2003;60:511–41. tivity in the motor cortex of resting human brain using echo-planar 8. Brodal P. The corticopontine projection in the rhesus monkey. MRI. Magn Res Med. 1995;34:537–41. Origin and principles of organization. Brain. 1978;101:251–83. 32. Beckmann CF, DeLuca M, Devlin JT, Smith SM. Investigations 9. Glickstein M, May JG, Mercier BE. Corticopontine projection in into resting-state connectivity using independent component analy- the macaque: the distribution of labeled cortical cells after large sis. Philos Trans R Soc Lond B Biol Sci. 2005;360:1001–13. injections of horseradish peroxidase in the pontine nuclei. J Comp 33. Krienen FM, Buckner RL. Segregated fronto-cerebellar circuits re- Neurol. 1985;235:343–59. vealed by intrinsic functional connectivity. Cereb Cortex. 2009;19: 10. Schmahmann JD, Pandya DN. Anatomical investigation of projec- 2485–97. tions to the basis pontis from posterior parietal association cortices 34. O’Reilly JX, Beckmann CF, Tomassini V, Ramnani N, Johansen-Berg in rhesus monkey. J Comp Neurol. 1989;289:53–73. H. Distinct and overlapping functional zones in the cerebellum defined 11. Schmahmann JD, Rosene DL, Pandya DN. Motor projections to the by resting state functional connectivity. Cereb Cortex. 2009;20:953–65. basis pontis in rhesus monkey. J Comp Neurol. 2004;478:248–68. 35. Habas C, Kamdar N, Nguyen D, Prater K, Beckmann CF, Menon V, 12. Schmahmann JD, MacMore J, Ko R. The human basis pontis. et al. Distinct cerebellar contributions to intrinsic connectivity net- Clinical syndromes and topographic organization. Brain. works. J Neurosci. 2009;29:8586–94. 2004;127:1269–91. 36. Stoodley CJ, Schmahmann JD. Evidence for topographic organiza- 13. Schmahmann JD, Pandya DN. Anatomic organization of the basilar tion in the cerebellum of motor control versus cognitive and affec- pontine projections from prefrontal cortices in rhesus monkey. J tive processing. Cortex. 2010;46:831–44. Neurosci. 1997;17:438–58. 216 Cerebellum (2015) 14:197–220 37. Sokolov AA, Erb M, Grodd W, Pavlova MA. Structural loop 62. Young JZ. Cerebellum and control of eye-movements in cephalo- between the cerebellum and the superior temporal sulcus: evidence pods. Nature. 1976;264:572–4. from diffusion tensor imaging. Cereb Cortex. 2014;24:626–32. 63. Bell CC, Han V, Sawtell NB. Cerebellum-like structures and their 38. Habas C, Guillevin R, Abanou A. Functional connectivity of the implications for cerebellar function. Annu Rev Neurosci. 2008;31: superior human temporal sulcus in the brain resting state at 3 T. 1–24. Neuroradiology. 2011;53:129–40. 64. Montgomery JC, Bodznick D, Yopak KE. The cerebellum and 39. Sang L, Qin W, Liu Y, Han W, Zhang Y, Jiang T, et al. Resting-state cerebellum-like structures of cartilaginous fishes. Brain Behav functional connectivity of the vermal and hemispheric subregions of Evol. 2012;80:152–65. cerebellum with both cerebral and cortical networks and subcortical 65. Oertel D, Young ED. What’s a cerebellar circuit doing in the structures. Neuroimage. 2012;61:1213–25. auditory system? Trends Neurosci. 2004;27:104–10. 40. Ding K, Liu Y, Yan X, Lin X, Jiang T. Altered functional connec- 66. Farris SM. Are mushroom bodies cerebellum-like structures? tivity of the primary visual cortex in subjects with amblyopia. Arthropod Struct Dev. 2011;40:368–79. Neural Plast. 2013;2013:612086. 67. Zeil J. Visual homing: an insect perspective. Curr Opin Neurobiol. 41. Seeley WW, Menon V, Schatzberg AF, Keller J, Glover GH, Kenna 2012;22:285–93. H, et al. Dissociable intrinsic connectivity networks for salience 68. Purcell EM. Life at low Reynolds number. Am J Phys. 1977;45:3– processing and executive control. J Neurosci. 2007;27:2349–56. 11. 42. Baumann O, Mattingley JB. Functional topography of primary 69. Gould SJ. The structure of evolutionary theory. Cambridge: emotion processing in the human cerebellum. Neuroimage. Belknap Press; 2002. 2012;61:805–11. 70. Brown RG, Hwang PYC. Introduction to random signals and ap- 43. Habas C, Cabanis EA. Dissociation of the neural networks recruited plied Kalman filtering. 3rd ed. New York: Wiley; 1996. during a haptic object-recognition task: complementary results with 71. Haykin S. Kalman filtering and neural networks. New York: Wiley; a tensorial independent component analysis. AJNR Am J 2001. Neuroradiol. 2008;29:1715–21. 72. Doucet A, De Freitas N, Gordon N. Sequential Monte Carlo 44. Holmes G. Clinical symptoms of cerebellar disease—and their methods in practice. New York: Springer; 2001. interpretation. Lancet. 1922;2:59–65. 73. Paulin MG. Evolution of the cerebellum as a neuronal ma- 45. Schmahmann J. The cerebellum and cognition. International review chine for Bayesian state estimation. J Neural Eng. 2005;2: of neurobiology, vol. 41. San Diego: Academic; 1997. 219–34. 46. Bullock TH. Electroreception. Ann Rev Neurosci. 1982;5:121–70. 74. Paulin MG, Hoffman LF. Bayesian head state prediction: computing 47. Llinas RR. Cortex of cerebellum. Sci Am. 1975;232:56–71. the dynamic prior with spiking neurons. Shanghai: International 48. Nieuwenhuys R, Pouwels E. A light and EM study of lobe C1 of the Conference on Natural Computation ICNC; 2011. gigantocerebellum of the mormyrid fish Gnathonemus petersi.Anat 75. Buesing L, Bill J, Nessler B, Maass W. Neural dynamics as sam- Rec. 1973;175:399. pling: a model for stochastic computation in recurrent networks of 49. Paulin MG. The role of the cerebellum in motor control and per- spiking neurons. PLoS Comput Biol. 2011;7:11. ception. Brain Behav Evol. 1993;41:39–50. 76. Deneve S. Making decisions with unknown sensory reliability. 50. Bower JM, Parsons LM. Rethinking the lesser brain. Sci Am. Front Neurosci. 2012;6:75. 2003;289:50–7. 77. Franklin DW, Wolpert DM. Computational mechanisms of senso- 51. Gao JH, Parsons LM, Bower JM, Xiong JH, Li JQ, Fox PT. rimotor control. Neuron. 2011;72:425–42. Cerebellum implicated in sensory acquisition and discrimination 78. Knill DC, Pouget A. The Bayesian brain: the role of uncertainty in rather than motor control. Science. 1996;272:545–7. neural coding and computation. Trends Neurosci. 2004;27:712–9. 52. Hore J, Watts S. Skilled throwers use physics to time ball release to 79. Kording KP, Wolpert DM. Bayesian integration in sensorimotor the nearest millisecond. J Neurophysiol. 2011;106:2024–33. learning. Nature. 2004;427:244–7. 53. Klier EM, Angelaki DE. Gaze Stabilization and the VOR. In: Koob 80. Laurens J, Droulez J. Bayesian processing of vestibular information. GF, Le Moal M, Thompson RF, editors. Encyclopedia of behavioral Biol Cybern. 2007;96:389–404. neuroscience. Oxford: Academic Press; 2010. p. 569–75. 81. Lochmann T, Deneve S. Neural processing as causal inference. Curr 54. Kurtzer I, Trautman P, Rasquinha RJ, Bhanpuri NH, Scott SH, Opin Neurobiol. 2011;21:774–81. Bastian AJ. Cerebellar damage diminishes long-latency responses 82. Ma WJ, Beck JM, Latham PE, Pouget A. Bayesian inference with to multijoint perturbations. J Neurophysiol. 2013;109:2228–41. probabilistic population codes. Nat Neurosci. 2006;9:1432–8. 55. Miall RC, King D. State estimation in the cerebellum. Cerebellum. 83. Wolpert DM. Probabilistic models in human sensorimotor control. 2008;7:572–6. Hum Mov Sci. 2007;26:511–24. 56. Molinari M, Restuccia D, Leggio MG. State estimation, response 84. Schmahmann JD. An emerging concept. The cerebellar contribution prediction, and cerebellar sensory processing for behavioral control. to higher function. Arch Neurol. 1991;48:1178–87. Cerebellum. 2009;8:399–402. 85. Snider RS, Stowell A. Receiving areas of the tactile, auditory, and 57. Parsons LM, Petacchi A, Schmahmann JD, Bower JM. Pitch dis- visual systems in the cerebellum. J Neurophysiol. 1944;7:331–57. crimination in cerebellar patients: evidence for a sensory deficit. 86. Ivry RB, Diener HC. Impaired velocity perception in patients with Brain Res. 2009;1303:84–96. lesions of the cerebellum. J Cogn Neurosci. 1991;3:355–66. 58. Roth MJ, Synofzik M, Lindner A. The cerebellum optimizes per- 87. Thier P, Haarmeier T, Treue S, Barash S. Absence of a common ceptual predictions about external sensory events. Curr Biol. functional denominator of visual disturbances in cerebellar disease. 2013;23:930–5. Brain. 1999;122:2133–46. 59. Zago M, McIntyre J, Senot P, Lacquaniti F. Visuo-motor coordina- 88. Ivry RB, Keele SW. Timing functions of the cerebellum. J Cogn tion and internal models for object interception. Exp Brain Res. Neurosci. 1989;1:136–52. 2009;192:571–604. 89. Baumann O, Mattingley JB. Scaling of neural responses to visual 60. Hobbs MJ, Young JZ. Cephalopod cerebellum. Brain Res. 1973;55: and auditory motion in the human cerebellum. J Neurosci. 2010;30: 424–30. 4489–95. 61. Williamson R, Chrachri A. A model biological neural network: the 90. Petacchi A, Kaernbach C, Ratnam R, Bower JM. Increased activa- cephalopod vestibular system. Philos Trans R Soc Lond B Biol Sci. tion of the human cerebellum during pitch discrimination: a positron 2007;362:473–81. emission tomography (PET) study. Hear Res. 2011;282:35–48. Cerebellum (2015) 14:197–220 217 91. Sokolov AA, Gharabaghi A, Tatagiba MS, Pavlova M. Cerebellar 113. Blake R, Turner LM, Smoski MJ, Pozdol SL, Stone WL. Visual engagement in an action observation network. Cereb Cortex. recognition of biological motion is impairedinchildrenwith autism. 2010;20:486–91. Psychol Sci. 2003;14:151–7. 92. Sokolov AA, Erb M, Gharabaghi A, Grodd W, Tatagiba MS, 114. Kim J, Doop ML, Blake R, Park S. Impaired visual recognition of Pavlova MA. Biological motion processing: the left cerebellum biological motion in schizophrenia. Schizophr Res. 2005;77:299– communicates with the right superior temporal sulcus. 307. Neuroimage. 2012;59:2824–30. 115. Barnea-Goraly N, Kwon H, Menon V, Eliez S, Lotspeich L, Reiss 93. Baumann O, Greenlee MW. Neural correlates of coherent audiovi- AL. White matter structure in autism: preliminary evidence from sual motion perception. Cereb Cortex. 2007;17:1433–43. diffusion tensor imaging. Biol Psychiatry. 2004;55:323–6. 94. Bushara KO, Grafman J, Hallett M. Neural correlates of auditory– 116. Kanaan RA, Borgwardt S, McGuire PK, Craig MC, Murphy DG, visual stimulus onset asynchrony detection. J Neurosci. 2001;21: Picchioni M, et al. Microstructural organization of cerebellar tracts 300–4. in schizophrenia. Biol Psychiatry. 2009;66:1067–9. 95. Calvert GA, Hansen PC, Iversen SD, Brammer MJ. Detection of 117. Kaiser MD, Hudac CM, Shultz S, Lee SM, Cheung C, Berken AM, audio-visual integration sites in humans by application of electro- et al. Neural signatures of autism. Proc Natl Acad Sci U S A. physiological criteria to the BOLD effect. Neuroimage. 2001;14: 2010;107:21223–8. 427–38. 118. Catani M, Jones DK, Daly E, Embiricos N, Deeley Q, Pugliese L, 96. Cerminara NL, Apps R, Marple-Horvat DE. An internal model of a et al. Altered cerebellar feedback projections in Asperger syndrome. moving visual target in the lateral cerebellum. J Physiol. 2009;587: Neuroimage. 2008;41:1184–91. 429–42. 119. Nawrot M, Rizzo M. Motion perception deficits from midline 97. Bower JM. Control of sensory data acquisition. In: Schmahmann cerebellar lesions in human. Vision Res. 1995;35:723–31. JD, editor. The cerebellum and cognition. International review of 120. Sokolov AA, Erb M, Grodd W, Tatagiba MS, Frackowiak RS, neurobiology, vol. 41. San Diego: Academic; 1997. p. 489–513. Pavlova MA. Recovery of biological motion perception and net- 98. Keele SW, Ivry R. Does the cerebellum provide a common compu- work plasticity after cerebellar tumor removal. Cortex. 2014. doi:10. tation for diverse tasks? A timing hypothesis. Ann N Y Acad Sci. 1016/j.cortex.2014.05.012. 1990;608:179–207. 121. Grèzes J, Frith CD, Passingham RE. Inferring false beliefs from the 99. Pavlova MA. Biological motion processing as a hallmark of social actions of oneself and others: an fMRI study. Neuroimage. 2004;21: cognition. Cereb Cortex. 2012;22:981–95. 744–50. 100. Johansson G. Visual perception of biological motion and a model 122. Ohnishi T, Moriguchi Y, Matsuda H, Mori T, Hirakata M, for its analysis. Percept Psychophys. 1973;14:201–11. Imabayashi E, et al. The neural network for the mirror system and 101. Runeson S, Frykholm G. Kinematic specification of dynamics as an mentalizinginnormallydeveloped children:anfMRIstudy. informational basis for person and action perception: expectation, Neuroreport. 2004;15:1483–7. gender recognition, and deceptive intention. J Exp Psychol. 123. Gobbini MI, Koralek AC, Bryan RE, Montgomery KJ, Haxby JV. 1983;112:585–615. Two takes on the social brain: a comparison of theory of mind tasks. 102. Saygin AP. Superior temporal and premotor brain areas necessary J Cogn Neurosci. 2007;19:1803–14. for biological motion perception. Brain. 2007;130:2452–61. 124. Jack A, Pelphrey KA. Neural correlates of animacy attribution 103. Bonda E, Petrides M, Ostry D, Evans A. Specific involvement of include neocerebellum in healthy adults. Cereb Cortex. 2014. doi: human parietal systems and the amygdala in the perception of 10.1093/cercor/bhu146. biological motion. J Neurosci. 1996;16:3737–44. 125. Jack A, Englander ZA, Morris JP. Subcortical contributions to 104. Vaina LM, Solomon J, Chowdhury S, Sinha P, Belliveau JW. effective connectivity in brain networks supporting imitation. Functional neuroanatomy of biological motion perception in Neuropsychologia. 2011;49:3689–98. humans. Proc Natl Acad Sci U S A. 2001;98:11656–61. 126. Cullen KE. The neural encoding of self-motion. Curr Opin 105. Pavlova M, Lutzenberger W, Sokolov A, Birbaumer N. Dissociable Neurobiol. 2011;21:587–95. cortical processing of recognizable and non-recognizable biological 127. Goldberg JM, Wilson VJ, Cullen KE, et al. The cerebellum and the movement: analysing gamma MEG activity. Cereb Cortex. vestibular system. In: Goldberg JM, Wislon VJ, Cullen KE, 2004;14:181–8. Angelaki DE, Broussard DM, Buttner-Ennever J, Fukushima K, 106. Oram MW, Perrett DI. Integration of form and motion in the anterior Minor LB, editors. The vestibular system, a sixth sense. Oxford: superior temporal polysensory area (STPa) of the macaque monkey. Oxford University Press; 2012. J Neurophysiol. 1996;76:109–29. 128. Bense S, Janusch B, Vucurevic G, Bauermann T, Schlindwein P, 107. Pelphrey KA, Mitchell TV, McKeown MJ, Goldstein J, Allison T, Brandt T, et al. Brainstem and cerebellar fMRI-activation during McCarthy G. Brain activity evoked by the perception of human horizontal and vertical optokinetic stimulation. Exp Brain Res. walking: controlling for meaningful coherent motion. J Neurosci. 2006;174:312–23. 2003;23:6819–25. 129. Kleinschmidt A, Thilo KV, Büchel C, Gresty MA, Bronstein AM, 108. Pavlova MA, Sokolov AN, Bidet-Ildei C. Sex differences in the Frackowiak RS. Neural correlates of visual-motion perception as neuromagnetic cortical response to body motion. Cereb Cortex. object- or self-motion. Neuroimage. 2002;16:873–82. 2014. doi:10.1093/cercor/bhu175. 130. Bronstein AM, Grunfeld EA, Faldon M, Okada T. Reduced self- 109. Grossman E, Donnelly M, Price R, Pickens D, Morgan V, Neighbor motion perception in patients with midline cerebellar lesions. G, et al. Brain areas involved in perception of biological motion. J Neuroreport. 2008;19:691–3. Cogn Neurosci. 2000;12:711–20. 131. Bertolini G, Ramat S, Bockisch CJ, Marti S, Straumann D, Palla A. 110. Strick PL, Dum RP, Fiez JA. Cerebellum and nonmotor function. Is vestibular self-motion perception controlled by the velocity stor- Annu Rev Neurosci. 2009;32:413–34. age? Insights from patients with chronic degeneration of the 111. Glickstein M, Gerrits N, Kralj-Hans I, Mercier B, Stein J, Voogd J. vestibulo-cerebellum. PLoS One. 2012;7:6. Visual pontocerebellar projections in the macaque. J Comp Neurol. 132. Merfeld DM, Zupan L, Peterka RJ. Humans use internal models to 1994;349:51–72. estimate gravity and linear acceleration. Nature. 1999;398:615–618. 112. Schmahmann JD, Pandya DN. Course of the fiber pathways to pons 133. Laurens J, Meng H, Angelaki DE. Neural representation of orien- from parasensory association areas in the rhesus monkey. J Comp tation relative to gravity in the macaque cerebellum. Neuron. Neurol. 1992;326:159–79. 2013;80:1508–18. 218 Cerebellum (2015) 14:197–220 134. Brooks JX, Cullen KE. Multimodal integration in rostral fastigial 156. Knolle F, Schraeger E, Baess P, Kotz SA. Cerebellar contributions nucleus provides an estimate of body movement. J Neurosci. to the prediction of self-initiated sounds. Cortex. 2013;49:2449–61. 2009;29:10499–511. 157. Friston K. The free-energy principle: a unified brain theory? Nat 135. Kleine JF, Guan Y, Kipiani E, Glonti L, Hoshi M, Buttner U. Trunk Rev Neurosci. 2010;11:127–38. position influences vestibular responses of fastigial nucleus neurons 158. Blakemore SJ, Frith CD, Wolpert DM. Spatio-temporal prediction in the alert monkey. J Neurophysiol. 2004;91:2090–100. modulates the perception of self-produced stimuli. J Cogn Neurosci. 136. Shaikh AG, Meng H, Angelaki DE. Multiple reference frames for 1999;11:551–9. motion in the primate cerebellum. J Neurosci. 2004;24:4491–7. 159. Kitazawa S, Kohno T, Uka T. Effects of delayed visual information 137. Brooks JX, Cullen KE. The primate cerebellum selectively encodes on the rate and amount of prism adaptation in the human. J unexpected self-motion. Curr Biol. 2013;23:947–55. Neurosci. 1995;15:7644–52. 138. Rochefort C, Arabo A, André M, Poucet B, Save E, Rondi-Reig L. 160. Tesche CD, Karhu JJ. Anticipatory cerebellar responses during Cerebellum shapes hippocampal spatial code. Science. 2011;334: somatosensory omission in man. Hum Brain Mapp. 2000;9:119–42. 385–9. 161. Wu X, Nestrasil I, Ashe J, Tuite P, Bushara K. Inferior olive 139. Batton 3rd RR, Jayaraman A, Ruggiero D, Carpenter MB. Fastigial response to passive tactile and visual stimulation with variable efferent projections in the monkey: an autoradiographic study. J interstimulus intervals. Cerebellum. 2010;9:598–602. Comp Neurol. 1977;174:281–305. 162. Moberget T, Karns CM, Deouell LY, Lindgren M, Knight RT, Ivry 140. Kamishina H, Conte WL, Patel SS, Tai RJ, Corwin JV, Reep RL. RB. Detecting violations of sensory expectancies following cere- Cortical connections of the rat lateral posterior thalamic nucleus. bellar degeneration: a mismatch negativity study. Brain Res. 2009;1264:39–56. Neuropsychologia. 2008;46:2569–79. 141. Moulton EA, Schmahmann JD, Becerra L, Borsook D. The cere- 163. Händel B, Thier P, Haarmeier T. Visual motion perception deficits bellum and pain: passive integrator or active participator? Brain Res due to cerebellar lesions are paralleled by specific changes in Rev. 2010;65:14–27. cerebro-cortical activity. J Neurosci. 2009;29:15126–33. 142. Saab CY, Willis WD. The cerebellum: organization, functions and 164. O’Reilly JX, Mesulam MM, Nobre AC. The cerebellum predicts the its role in nociception. Brain Res Rev. 2003;42:85–95. timing of perceptual events. J Neurosci. 2008;28:2252–60. 143. Ekerot CF, Garwicz M, Schouenborg J. The postsynaptic dorsal 165. Grube M, Cooper FE, Chinnery PF, Griffiths TD. Dissociation of column pathway mediates cutaneous nociceptive information to duration-based and beat-based auditory timing in cerebellar degen- cerebellar climbing fibres in the cat. J Physiol. 1991;441:275–84. eration. Proc Natl Acad Sci U S A. 2010;107:11597–601. 144. Wu J, Chen PX. Discharge response of cerebellar Purkinje cells to 166. Koch G, Oliveri M, Caltagirone C. Neural networks engaged in stimulation of C-fiber in cat saphenous nerve. Brain Res. 1992;581: milliseconds and seconds time processing: evidence from transcra- 269–72. nial magnetic stimulation and patients with cortical or subcortical 145. Apkarian AV, Bushnell MC, Treede RD, Zubieta JK. Human brain dysfunction. Philos Trans R Soc Lond B Biol Sci. 2009;364:1907– mechanisms of pain perception and regulation in health and disease. 18. Eur J Pain. 2005;9:463–84. 167. Grube M, Lee KH, Griffiths TD, Barker AT, Woodruff PW. 146. Helmchen C, Mohr C, Erdmann C, Binkofski F. Cerebellar neural Transcranial magnetic theta-burst stimulation of the human cerebel- responses related to actively and passively applied noxious thermal lum distinguishes absolute, duration-based from relative, beat- stimulation in human subjects: a parametric fMRI study. Neurosci based perception of subsecond time intervals. Front Psychol. Lett. 2004;361:237–40. 2010;1:171. 147. Borsook D, Moulton EA, Tully S, Schmahmann JD, Becerra L. 168. Harrington DL, Lee RR, Boyd LA, Rapcsak SZ, Knight RT. Does Human cerebellar responses to brush and heat stimuli in healthy and the representation of time depend on the cerebellum? Effect of neuropathic pain subjects. Cerebellum. 2008;7:252–72. cerebellar stroke. Brain. 2004;127:561–74. 148. Moulton EA, Elman I, Pendse G, Schmahmann J, Becerra L, 169. Gooch CM, Wiener M, Wencil EB, Coslett HB. Interval timing Borsook D. Aversion-related circuitry in the cerebellum: responses disruptions in subjects with cerebellar lesions. Neuropsychologia. to noxious heat and unpleasant images. J Neurosci. 2011;31:3795– 2010;48:1022–31. 804. 170. Ivry RB, Spencer RM. Evaluating the role of the cerebellum in 149. Moulton EA, Elman I, Becerra LR, Goldstein RZ, Borsook D. The temporal processing: beware of the null hypothesis. Brain. cerebellum and addiction: insights gained from neuroimaging re- 2004;127:E13. search. Addict Biol. 2014;19:317–31. 171. Lewis PA, Miall RC. Brain activation patterns during measurement 150. Thompson RF. Neural mechanisms of classical conditioning in of sub- and supra-second intervals. Neuropsychologia. 2003;41: mammals. Philos Trans R Soc Lond B Biol Sci. 1990;329:161–70. 1583–92. 151. Koekkoek SK, Hulscher HC, Dortland BR, Hensbroek RA, 172. Bueti D, Lasaponara S, Cercignani M, Macaluso E. Learning about Elgersma Y, Ruigrok TJ, et al. Cerebellar LTD and learning- time: plastic changes and interindividual brain differences. Neuron. dependent timing of conditioned eyelid responses. Science. 2012;75:725–37. 2003;301:1736–9. 173. Wiener M, Lee YS, Lohoff FW, Coslett HB. Individual differences 152. Perrett SP, Ruiz BP, Mauk MD. Cerebellar cortex lesions disrupt in the morphometry and activation of time perception networks are learning-dependent timing of conditioned eyelid responses. J influenced by dopamine genotype. Neuroimage. 2014;89:10–22. Neurosci. 1993;13:1708–18. 174. Hayashi MJ, Kantele M, Walsh V, Carlson S, Kanai R. Dissociable 153. Bares M, Lungu O, Liu T, Waechter T, Gomez CM, Ashe J. neuroanatomical correlates of subsecond and suprasecond time Impaired predictive motor timing in patients with cerebellar disor- perception. J Cogn Neurosci. 2014;26:1685–93. ders. Exp Brain Res. 2007;180:355–65. 175. Paquette S, Mignault Goulet G, Rothermich K. Prediction, attention 154. Rahmati N, Owens CB, Bosman LW, Spanke JK, Lindeman S, and unconscious processing in hierarchical auditory perception. Gong W, et al. Cerebellar potentiation and learning a whisker- Front Psychol. 2013;4:955–6. based object localization task with a time response window. J 176. Clark A. Whatever next? Predictive brains, situated agents, and the Neurosci. 2014;34:1949–62. future of cognitive science. Behav Brain Sci. 2013;36:181–204. 155. Shadmehr R, Smith MA, Krakauer JW. Error correction, sensory 177. Popa LS, Hewitt AL, Ebner TJ. Predictive and feedback perfor- prediction, and adaptation in motor control. Annu Rev Neurosci. mance errors are signaled in the simple spike discharge of individual 2010;33:89–108. Purkinje cells. J Neurosci. 2012;32:15345–58. Cerebellum (2015) 14:197–220 219 178. Molinari M, Chiricozzi F, Clausi S, Tedesco A, De Lisa M, Leggio 200. Fatemi SH, Aldinger KA, Ashwood P, et al. Consensus paper: M. Cerebellum and detection of sequences, from perception to pathological role of the cerebellum in autism. Cerebellum. cognition. Cerebellum. 2008;7:611–5. 2012;11:777–807. 179. Braitenberg V, Heck D, Sultan F. The detection and generation of 201. Larson JCG, Mostofsky SH. Evidence that the pattern of sequences as a key to cerebellar function: experiments and theory. visuomotor sequence learning is altered in children with autism. Behav Brain Sci. 1997;20:229–77. Autism Res. 2008;1:341–53. 180. Ivry R. Exploring the role of the cerebellum in sensory anticipation 202. Remillard G. The study of sequence learning in individuals with and timing: commentary on Tesche and Karhu. Hum Brain Mapp. schizophrenia: a critical review of the literature. J Neuropsychol. 2000;9:115–8. 2014;8:231–45. 181. Nixon PD. The role of the cerebellum in preparing responses to 203. Bower JM. Is the cerebellum sensory for motor’s sake, of motor for predictable sensory events. Cerebellum. 2003;2:114–22. sensory’s sake: the view from the whiskers of a rat? Prog Brain Res. 182. Restuccia D, Della MG, Valeriani M, Leggio MG, Molinari M. 1997;114:483–516. Cerebellar damage impairs detection of somatosensory input chang- 204. Bower JM, Kassel J. Variability in tactile projection patterns to es. A somatosensory mismatch-negativity study. Brain. 2007;130: cerebellar folia crus IIa of the Norway rat. J Comp Neurol. 276–87. 1990;302:768–78. 183. Ito M. Cerebellar circuitry as a neuronal machine. Prog Neurobiol. 205. Bower JM. Computational structure of the cerebellar molecular 2006;78:272–303. layer. In: Manto M, Gruol D, Schmahmann J, Koibuchi N, Rossi 184. Molinari M, Leggio MG, Solida A, Ciorra R, Misciagna S, Silveri F, editors. Handbook of the cerebellum and cerebellar disorders. MC, et al. Cerebellum and procedural learning: evidence from focal New York: Springer; 2013. p. 1359–80. cerebellar lesions. Brain. 1997;120:1753–62. 206. Parsons LM, Bower JM, Gao JH, Xiong J, Li J, Fox PT. Lateral 185. Leggio MG, Chiricozzi FR, Clausi S, Tedesco AM, Molinari M. cerebellar hemispheres actively support sensory acquisition and The neuropsychological profile of cerebellar damage: the sequenc- discrimination rather than motor control. Learn Mem. 1997;4:49– ing hypothesis. Cortex. 2011;47:137–44. 62. 186. Dirnberger G, Novak J, Nasel C, Zehnter M. Separating coordina- 207. Liu Y, Pu Y, Gao JH, Parsons LM, Xiong J, Liotti M, et al. The tive and executive dysfunction in cerebellar patients during motor human red nucleus and lateral cerebellum in supporting roles for skill acquisition. Neuropsychologia. 2010;48:1200–8. sensory information processing. Hum Brain Mapp. 2000;10:147– 187. Doyon J, Gaudreau D, Laforce RJ, Castonguay M, Bedard PJ, 59. Bedard F, et al. Role of the striatum, cerebellum, and frontal lobes 208. Petacchi A, Laird AR, Fox PT, Bower JM. Cerebellum and auditory in the learning of a visuomotor sequence. Brain Cogn. 1997;34: function: an ALE meta-analysis of functional neuroimaging studies. 218–45. Hum Brain Mapp. 2005;25:118–228. 188. Ferrucci R, Brunoni A, Parazzini M, Vergari M, Rossi E, Fumagalli 209. Kandel ER, Schwartz JH, Jessel TM. Principles of neural science. M, et al. Modulating human procedural learning by cerebellar Norwalk: Appleton and Lange; 1991. transcranial direct current stimulation. Cerebellum. 2013;12:485– 210. Tinazzi M, Morgante F, Peretti A, Mariotti C, Panzeri M, Fiorio M, 92. et al. Impaired temporal processing of tactile and proprioceptive 189. Gomez-Beldarrain M, Garcia-Monco JC, Rubio B, Pascual-Leone stimuli in cerebellar degeneration. PLoS One. 2013;8:e78628. A. Effect of focal cerebellar lesions on procedural learning in the 211. Bhanpuri NH, Okamura AM, Bastian AJ. Active force perception serial reaction time task. Exp Brain Res. 1998;120:25–30. depends on cerebellar function. J Neurophysiol. 2012;107:1612–20. 190. Nixon PD, Passingham RE. The cerebellum and cognition: cerebel- 212. Christensen A, Giese MA, Sultan F, Mueller OM, Goericke SL, Ilg lar lesions impair sequence learning but not conditional visuomotor W, et al. An intact action–perception coupling depends on the learning in monkeys. Neuropsychologia. 2000;38:1054–72. integrity of the cerebellum. J Neurosci. 2014;34:6707–16. 191. Timmann D, Drepper J, Calabrese S, Burgerhoff K, Maschke M, 213. Ackermann H, Mathiak K, Riecker A. The contribution of the Kolb FP, et al. Use of sequence information in associative learning cerebellum to speech production and speech perception: clinical in control subjects and cerebellar patients. Cerebellum. 2004;3:75– and functional imaging data. Cerebellum. 2007;6:202–13. 82. 214. Jokisch D, Troje NF, Koch B, Schwarz M, Daum I. Differential 192. Leggio MG, Tedesco AM, Chiricozzi FR, Clausi S, Orsini A, involvement of the cerebellum in biological and coherent motion Molinari M. Cognitive sequencing impairment in patients with focal perception. Eur J Neurosci. 2005;21:3439–46. or atrophic cerebellar damage. Brain. 2008;13:1332–43. 215. Lee KH, Egleston PN, Brown WH, Gregory AN, Barker AT, 193. Gomot M, Wicker B. A challenging, unpredictable world for people Woodruff PW. The role of the cerebellum in subsecond time per- with autism spectrum disorder. Int J Psychophysiol. 2012;83:240–7. ception: evidence from repetitive transcranial magnetic stimulation. 194. Ford JM, Mathalon DH. Anticipating the future: automatic predic- J Cogn Neurosci. 2007;19:147–57. tion failures in schizophrenia. Int J Psychophysiol. 2012;83:232–9. 216. Bower JM. Functional implications of tactile projection patterns to 195. Horga G, Schatz KC, Abi-Dargham A, Peterson BS. Deficits in the lateral hemispheres of the cerebellum of the albino rat: the legacy predictive coding underlie hallucinations in schizophrenia. J of Wally Welker. Ann N Y Acad Sci. 2011;1225:130–41. Neurosci. 2014;34:8072–82. 217. Proville RD, Spolidoro M, Guyon N, Dugue GP, Selimi F, Isope P, 196. Shergill SS, White TP, Joyce DW, Bays PM, Wolpert DM, Frith CD. et al. Cerebellum involvement in cortical sensorimotor circuits for the Functional magnetic resonance imaging of impaired sensory pre- control of voluntary movements. Nat Neurosci. 2014;17:1233–9. diction in schizophrenia. JAMA Psychiatry. 2014;71:28–35. 218. Flourens P. Recherches Experimentales Sur Les Proprietes Et Les 197. Jack A, Morris JP. Neocerebellar contributions to social perception Fonctions Du Systeme Nerveux, Dans Les Animaux Vertebres. in adolescents with autism spectrum disorder. Dev Cogn Neurosci. Crevot: Paris; 1824. 2014;10:77–92. 219. Holmes G. The cerebellum of man. Brain. 1939;62:1–30. 198. Sachse M, Schlitt S, Hainz D, Ciaramidaro A, Walter H, Poustka F, 220. Manto M, Bower JM, Conforto AB, Delgado-Garcia JM, da Guarda et al. Facial emotion recognition in paranoid schizophrenia and SN, Gerwig M, et al. Consensus paper: roles of the cerebellum in autism spectrum disorder. Schizophr Res. 2014. doi:10.1016/j. motor control—the diversity of ideas on cerebellar involvement in schres.2014.08.030. movement. Cerebellum. 2012;11(2):457–87. 199. Andreasen NC, Pierson R. The role of the cerebellum in schizo- 221. Schlerf JE, Xu J, Klemfuss NM, Griffiths TL, Ivry RB. Individuals phrenia. Biol Psychiatry. 2008;64:81–8. with cerebellar degeneration show similar adaptation deficits with 220 Cerebellum (2015) 14:197–220 large and small visuomotor errors. J Neurophysiol. 2013;109:1164– 234. Kellermann T, Regenbogen C, De Vos M, Mößnang C, Finkelmeyer 73. A, Habel U. Effective connectivity of the human cerebellum during 222. Thompson RF, Steinmetz JE. The role of the cerebellum in classical visual attention. J Neurosci. 2012;32:11453–60. conditioning of discrete behavioral responses. Neuroscience. 235. Ravizza SM, McCormick CA, Schlerf JE, Justus T, Ivry RB, Fiez 2009;162:732–55. JA. Cerebellar damage produces selective deficits in verbal working 223. Huang C, Ricklefs RE. Cerebellum and human evolution: a com- memory. Brain. 2006;129:306–20. parative and information theory perspective. In: Manto M, Gruol D, 236. Robinson DA. The use of control systems analysis in the neuro- Schmahmann J, Koibuchi N, Rossi F, editors. Handbook of the physiology of eye movements. Annu Rev Neurosci. 1981;4:463– cerebellum and cerebellar disorders. New York: Springer; 2013. p. 503. 1337–57. 237. Lisberger SG. Neural basis for motor learning in the vestibuloocular 224. Kawashima R, Okuda J, Umetsu A, Sugiura M, Inoue K, Suzuki K, reflex of primates. III. Computational and behavioral analysis of the et al. Human cerebellum plays an important role in memory-timed sites of learning. J Neurophysiol. 1994;72:974–98. finger movement: an fMRI study. J Neurophysiol. 2000;83:1079–87. 238. Westheimer G, McKee SP. Visual acuity in the presence of retinal- 225. D’Angelo E, Casali S. Seeking a unified framework for cerebellar image motion. J Opt Soc Am. 1975;65:847–50. function and dysfunction: from circuit operations to cognition. Front 239. Little WJ. On the influence of abnormal parturition, difficult la- Neural Circ. 2012;6:116. bours, premature birth and asphxia neonatorum on the mental and 226. Schlerf J, Ivry RB, Diedrichsen J. Encoding of sensory physical condition of the child, especially in relation to deformities. prediction errors in the human cerebellum. J Neurosci. Clin Orthop Relat Res. 1966;46:7–22. 2012;32:4913–22. 240. Spence RD, Zhen Y, White S, Schlinger BA, Day LB. Recovery of 227. Ivry RB, Spencer RMC. The neural representation of time. Curr motor and cognitive function after cerebellar lesions in a songbird: Opin Neurobiol. 2004;14:225–32. role of estrogens. Eur J Neurosci. 2009;29:1225–34. 228. Holscher C, Munk M. Information processing by neuronal popula- 241. Williams AJ, Ling GSF, Tortella FC. Severity level and injury track tions. Cambridge: Cambridge University Press; 2008. determine outcome following a penetrating ballistic-like brain inju- 229. Frings M, Boenisch R, Gerwig M, Diener HC, Timmann D. ry in the rat. Neurosci Lett. 2006;408:183–8. Learning of sensory sequences in cerebellar patients. Learn Mem. 242. Willson ML, Bower AJ, Sherrard RM. Developmental neural plas- 2004;11:347–55. ticity and its cognitive benefits: olivocerebellar reinnervation com- 230. Pfordresher PQ, Mantell JT, Brown S, Zivadinov R, Cox JL. pensates for spatial function in the cerebellum. Eur J Neurosci. Brain responses to altered auditory feedback during musical 2007;25:1475–83. keyboard production: an fMRI study. Brain Res. 2014;1556: 243. Konczak J, Schoch B, Dimitrova A, Gizewski E, Timmann D. 28–37. Functional recovery of children and adolescents after cerebellar 231. Bell C, Bodznick D, Montgomery J, Bastian J. The generation and tumour resection. Brain. 2005;128:1428–41. subtraction of sensory expectations within cerebellum-like struc- 244. Fatemi SH. Cerebellum and autism. Cerebellum. 2013;12:778–9. tures. Brain Behav Evol. 1997;50:17–31. 245. Iwanaga R, Honda S, Nakane H, Tanaka K, Toeda H, Tanaka G. 232. Dancey E, Murphy B, Srbely J, Yielder P. The effect of experimental Pilot study: efficacy of sensory integration therapy for Japanese pain on motor training performance and sensorimotor integration. children with high-functioning autism spectrum disorder. Occup Exp Brain Res. 2014;232:2879–89. Ther Int. 2014;21:4–11. 233. Gottwald B, Mihajlovic Z, Wilde B, Mehdorn HM. Does the 246. Pavlova M, Krageloh-Mann I, Sokolov A, Birbaumer N. cerebellum contribute to specific aspects of attention? Recognition of point-light biological motion displays by Neuropsychologia. 2003;41:1452–60. young children. Perception. 2001;30:925–33.

Journal

The Cerebellum – Unpaywall

Published: Dec 6, 2014

There are no references for this article.