Functional topography in the human cerebellum: A meta-analysis of neuroimaging studies
Introduction
The cerebellum is involved in a wide range of tasks, including sensorimotor control, language, spatial and executive functions. Deficits resulting from cerebellar lesions include motor dysmetria, ataxia, and intention tremor (Holmes, 1939), but also the cerebellar cognitive affective syndrome (Schmahmann and Sherman, 1998) including executive, visual–spatial, linguistic and emotional deficits, and even mutism and psychosis (Botez-Marquard et al., 1994, Grafman et al., 1992, Heath et al., 1979, Levisohn et al., 2000, Molinari et al., 2004, Rapoport et al., 2000, Riva and Giorgi, 2000, Schmahmann et al., 2007, Steinlin et al., 2003).
The anatomical basis of this proposed cerebellar role in non-motor function is the existence of cerebro-cerebellar channels (cortico-ponto-cerebellar and cerebello-thalamo-cortical loops) that link the cerebellum with motor cortices as well as with association cortices and paralimbic regions of the cerebral hemisphere (Botez et al., 1985, Kelly and Strick, 2003, Leiner et al., 1986, Middleton and Strick, 1994, Schmahmann, 1991, Schmahmann, 1996, Schmahmann and Pandya, 1989, Schmahmann and Pandya, 1997, Voogd and Glickstein, 1998). Experimental investigations in animals (Chambers and Sprague, 1955a, Chambers and Sprague, 1955b, Snider and Eldred, 1951), imaging studies in humans (Bushara et al., 2001, Grodd et al., 2001, Grodd et al., 2005) and clinical reports (e.g., Victor et al., 1959, Schoch et al., 2006) have supported the original hypothesis of Bolk (1906) that there is topography of motor function within the cerebellum (see Manni and Petrosini (2004) for a review). Contemporary clinical studies suggest that whereas the cerebellar anterior lobe is principally engaged in motor control, the cerebellar vermis is involved in affective processing, and the posterior cerebellum contributes to complex cognitive operations (Exner et al., 2004, Levisohn et al., 2000, Schmahmann, 2004, Schmahmann, 2007, Schmahmann and Sherman, 1998, Schoch et al., 2006, Tavano et al., 2007). Furthermore, consistent with the crossed cerebro-cerebellar fiber pathways, linguistic impairments can arise following right cerebellar hemisphere lesions, whereas visual–spatial difficulties may follow left cerebellar hemisphere damage (Fiez et al., 1992, Gottwald et al., 2004, Gross-Tsur et al., 2006, Hokkanen et al., 2006, Riva and Giorgi, 2000, Scott et al., 2001). Yet some studies fail to detect non-motor problems after cerebellar tumor or stroke, and others find no reliable structure–function relationships (see Frank et al., 2007). Better understanding of the functional topography of the cerebellum would enable these contradictory results to be clarified with respect to the location of cerebellar damage.
Positron emission tomography (PET) scans first demonstrated cerebellar activation during extremity and eye movements (Fox et al., 1985), and then revealed cerebellar activation in language studies (Petersen et al., 1988, Raichle et al., 1994). Subsequent evaluations showed cerebellar activation in tasks of sensory processing (Gao et al., 1996), appreciation of timed intervals (Jueptner et al., 1995), anticipatory planning and prediction (shifting attention tasks) (Allen et al., 1997), verbal working memory (Desmond et al., 1997), classical conditioning (Logan and Grafton, 1995), and mental imagery (Ryding et al., 1993), among others. There is now a plethora of studies using PET and functional magnetic resonance imaging (fMRI) in which cerebellar activation is noted during sensorimotor, cognitive, and emotional processing paradigms. There are no published fMRI reports, however, examining multiple behavioral domains within single individuals. We therefore performed a meta-analysis of published functional imaging studies to test our hypothesis that movement, cognition, and affective processing are topographically arranged within the cerebellum (Schmahmann, 1991, Schmahmann, 1996, Schmahmann, 2004). We predicted that this approach would provide insights into whether different aspects of sensorimotor function, cognition, and emotional processing activate geographically distinct cerebellar regions. This meta-analysis is facilitated by the recently introduced GingerALE software (Laird et al., 2005) that enables the calculation of activation likelihood estimates for each voxel in the brain (Turkeltaub et al., 2002), and by the MRI Atlas of the Human Cerebellum (Schmahmann et al., 2000), which together make it possible to map sites of activation from multiple imaging studies onto a single representative cerebellum.
Section snippets
Literature review
Articles were identified through a PubMed (http://www.ncbi.nlm.nih.gov/pubmed) search of “cerebell⁎ AND imaging” with the limits “Humans”, “English”, and “Adult 19–44 years”, which yielded 1118 articles. Of these, we eliminated studies that did not involve functional neuroimaging (e.g., structural and morphometric studies); those that did not report cerebellar activation or the coordinates of the activation in standard units (Montreal Neurological Institute [MNI; Collins et al., 1998] or
Peak coordinates
Table 3 shows the cluster sizes, weighted centers, peak coordinates and cerebellar lobules of the significant ALE maxima for each category of task. Fig. 1 shows the thresholded ALE activation maps for four representative rostral–caudal coronal sections. A rostral–caudal dimension to the activation patterns is evident, as is medial–lateral and lobular organization.
Sensorimotor tasks
Motor and somatosensory representations show largely overlapping activation patterns, with the major cluster focused in lobule V and
Discussion
This activation likelihood estimate (ALE) meta-analysis provides a quantitative summary of patterns of cerebellar activation found in healthy adults for tasks in which different types of information are being processed. The results indicate that there is a functional topography of the cerebellum, particularly with regard to sensorimotor vs. language, spatial, and working memory tasks. In agreement with known cerebellar homunculi (Snider and Eldred, 1951), sensorimotor tasks showed significant
Conclusion
The results of this meta-analysis show that different regions of the cerebellum process information from different domains. Our findings provide support for anatomical, physiological and clinical studies reporting the existence of sensorimotor (anterior lobe, lobule VIII), cognitive (posterior lobe, particularly lobules VI and VII [Crus I and Crus II]), and emotional (posterior vermis) regions of the cerebellum. The areas of overlap, and the inherent limitations of the meta-analysis approach,
Acknowledgments
Grant support was provided by the Birmingham Foundation, and a Fund for Medical Discovery fellowship award to C. Stoodley from the MGH Executive Committee on Research. We would like to thank Jason MacMore and Joseph DiNunzio for their technical assistance, and acknowledge the work of Amy Hurwitz, Russell Loeber and Jennifer Marjani on a previous version of this project.
References (160)
- et al.
Neuroanatomic observations of the brain in autism: a review and future directions
Int. J. Dev. Neurosci.
(2005) - et al.
Dissociable networks for the expectancy and perception of emotional stimuli in the human brain
NeuroImage
(2006) - et al.
The cerebellum and decision making under uncertainty
Brain. Res. Cogn. Brain. Res.
(2004) Control of sensory data acquisition
The pontocerebellar projection in the rhesus monkey: an experimental study with retrograde axonal transport of horseradish peroxidase
Neuroscience
(1979)Principles of organization of the corticopontocerebellar projection to crus II in the cat with particular reference to the parietal cortical areas
Neuroscience
(1983)- et al.
The corticopontocerebellar pathway to crus I in the cat as studied with anterograde and retrograde transport of horseradish peroxidase
Brain Res.
(1983) - et al.
Salient anatomic features of the cortico-ponto-cerebellar pathway
Prog. Brain Res.
(1997) - et al.
The influence of working memory load on phase specific patterns of cortical activity
Brain Res. Cogn. Brain Res.
(2004) - et al.
Cerebrocerebellar networks during articulatory rehearsal and verbal working memory tasks
NeuroImage
(2005)