The deep cerebellar nuclei (DCN) are a key element of the cortico-cerebellar loop. Because of their small size and functional diversity, it is difficult to study them using magnetic resonance imaging (MRI). To overcome these difficulties, we present here three related methodological advances. First, we used susceptibility-weighted imaging (SWI) at a high-field strength (7T) to identify the dentate, globose, emboliform and fastigial nucleus in 23 human participants. Due to their high iron content, the DCN are visible as hypo-intensities. Secondly, we generated probabilistic maps of the deep cerebellar nuclei in MNI space using a number of common normalization techniques. These maps can serve as a guide to the average location of the DCN, and are integrated into an existing probabilistic atlas of the human cerebellum (Diedrichsen et al., 2009). The maps also quantify the variability of the anatomical location of the deep cerebellar nuclei after normalization. Our results indicate that existing normalization techniques do not provide satisfactory overlap to analyze the functional specialization within the DCN. We therefore thirdly propose a ROI-driven normalization technique that utilizes both information from a T1-weighted image and the hypo-intensity from a T2*-weighted or SWI image to ensure overlap of the nuclei. These techniques will promote the study of the functional specialization of subregions of the DCN using MRI.

In Friedreich's ataxia (FA) the genetically decreased expression of the mitochondrial protein frataxin leads to disturbance of the mitochondrial iron metabolism. Within the cerebellum the dentate nuclei (DN) are primarily affected. Histopathological studies show atrophy and accumulation of mitochondrial iron in DN. Dentate iron content has been suggested as a biomarker to measure the effects of siderophores/antioxidant treatment of FA. We assessed the iron content and the volume of DN in FA patients and controls based on ultra-high-field MRI (7 Tesla) images.

The aim of the present study was to compare possible activation of the interposed and dentate cerebellar nuclei during finger, foot and tongue movements using functional magnetic resonance imaging (fMRI). Nineteen healthy control subjects performed sequential finger and repetitive tongue and foot movement tasks. Thin slices (2.5mm) were acquired of the cerebellar region containing the cerebellar nuclei with high spatial resolution (matrix size 128 x 128 x 10) using a Siemens 1.5T Sonata system. Use of an eight channel head coil provided better signal-to-noise-ratio compared to standard head coils. Only data of those 12 subjects were included in final statistical analysis, who showed significant activation of the cerebellar nuclei at least in one task. Cortical activations of the superior cerebellum were found in accordance to the known somatotopy of the human cerebellar cortex. Nuclear activations were most significant in the sequential finger movement task. Both interposed nuclei and ipsilateral dentate nucleus were activated. Dentate activation was present in the more caudal parts of both the dorsal and ventral nucleus. Activation overlapped with motor and non-motor domains of the dentate nucleus described by Dum and Strick [R.P. Dum, P.L. Strick, An unfolded map of the cerebellar dentate nucleus and its projections to the cerebral cortex, J. Neurophysiol. 89 (2003) 634-639] based on anatomical data in monkey. Tongue movement related activations were less extensive and overlapped with activations of caudal parts of the dentate nucleus in the finger movement task. No nuclear activation was seen following foot movements. The present findings show that both interposed and dentate nuclei are involved in sequential finger movements in humans. Interposed nucleus likely contributes to movement performance. Although no direct conclusions could be drawn based on the present data, different parts of the dentate nucleus may contribute to movement performance, planning and possible non-motor parts of the task.

The first aim of the present study was to extend previous findings of similar cerebellar cortical areas being involved in verbal and spatial n-back working memory to the level of the cerebellar nuclei. The second aim was to investigate whether different areas of the cerebellar cortex and nuclei contribute to different working memory tasks (n-back vs. Sternberg tasks). Young and healthy subjects participated in two functional magnetic resonance imaging (fMRI) studies using a 7 T MR scanner with its increased signal-to-noise ratio. One group of subjects (n=21) performed an abstract and a verbal version of an n-back task contrasting a 2-back and 0-back condition. Another group of subjects (n=23) performed an abstract and a verbal version of a Sternberg task contrasting a high load and a low load condition. A block design was used. For image processing of the dentate nuclei, a recently developed region of interest (ROI) driven normalization method of the dentate nuclei was applied (Diedrichsen et al., 2011). Whereas activated areas of the cerebellar cortex and dentate nuclei were not significantly different comparing the abstract and verbal versions of the n-back task, activation in the abstract and verbal Sternberg tasks was significantly different. In both n-back tasks activation was most prominent at the border of lobules VI and Crus I, within lobule VII, and within the more caudal parts of the dentate nucleus bilaterally. In Sternberg tasks the most prominent activations were found in lobule VI extending into Crus I on the right. In the verbal Sternberg task activation was significantly larger within right lobule VI compared to the abstract Sternberg task and compared to the verbal n-back task. Activations of rostral parts of the dentate were most prominent in the verbal Sternberg task, whereas activation of caudal parts predominated in the abstract Sternberg task. On the one hand, the lack of difference between abstract and verbal n-back tasks and the lack of significant lateralization suggest a more general contribution of the cerebellum to working memory regardless of the modality. On the other hand, the focus of activation in right lobule VI in the verbal Sternberg task suggests specific cerebellar contributions to verbal working memory. The verbal Sternberg task emphasizes maintenance of stimuli via phonological rehearsal, whereas central executive demands prevail in n-back tasks. Based on the model of working memory by Baddeley and Hitch (1974), the present results show that different regions of the cerebellum support functions of the central executive system and one of the subsidiary systems, the phonological loop.

Human cerebellar lesion studies provide good evidence that the cerebellum contributes to the acquisition of classically conditioned eyeblink responses (CRs). As yet, only one study used more advanced methods of lesion-symptom (or lesion-behavior) mapping to investigate which cerebellar areas are involved in CR acquisition in humans. Likewise, comparatively few studies investigated the contribution of the human cerebellum to CR extinction and savings. In this present study, young adults with focal cerebellar disease were tested. A subset of participants was expected to acquire enough conditioned responses to allow the investigation of extinction and saving effects. 19 participants with chronic surgical lesions of the cerebellum and 19 matched control subjects were tested. In all cerebellar subjects benign tumors of the cerebellum had been surgically removed. Eyeblink conditioning was performed using a standard short delay protocol. An initial unpaired control phase was followed by an acquisition phase, an extinction phase and a subsequent reacquisition phase. Structural 3T magnetic resonance images of the brain were acquired on the day of testing. Cerebellar lesions were normalized using methods optimized for the cerebellum. Subtraction analysis and Liebermeister tests were used to perform lesion-symptom mapping. As expected, CR acquisition was significantly reduced in cerebellar subjects compared to controls. Reduced CR acquisition was significantly more likely in participants with lesions of lobule VI and Crus I extending into Crus II (p<0.05, Liebermeister test). Cerebellar subjects could be subdivided into two groups: a smaller group (n=5) which showed acquisition, extinction and savings within the normal range; and a larger group (n=14) which did not show acquisition. In the latter, no conclusions on extinction or savings could be drawn. Previous findings were confirmed that circumscribed areas in lobule VI and Crus I are of major importance in CR acquisition. In addition, the present data suggest that if the critical regions of the cerebellar cortex are lesioned, the ability to acquire CRs is not only reduced but abolished. Subjects with lesions outside these critical areas, on the other hand show preserved acquisition, extinction and saving effects. As a consequence, studies in human subjects with cerebellar lesions do not allow drawing conclusions on CR extinction and savings. In light of the present findings, previous reports of reduced extinction in humans with circumscribed cerebellar disease need to be critically reevaluated.

Acquisition of conditioned eyeblink responses is known to decline with age, and age-related decline has been related to a reduction of cerebellar size and function. The aim of the present study was to investigate age-related effects on storage-related processes and extinction of visual threat eyeblink responses (VTERs), conditioned responses which are naturally acquired in early childhood. Storage and extinction of VTERs were tested in 34 healthy participants with an age range from 21 to 74 years (mean age 41.6±16.3 years). High-resolution structural magnetic resonance images (MRI) were acquired in all subjects. Conventional volumetric measures and voxel-based morphometry (VBM) were performed at the level of the cerebellum. Storage and extinction of VTERs showed a significant age-dependent decline. Likewise, cerebellar volume decreased with age. Storage, but not extinction showed a significant positive correlation with age-dependent reduction of total cerebellar volume. VBM analysis showed that gray matter volume in circumscribed areas of intermediate lobules VI, and Crus I and II bilaterally were positively correlated with VTER storage (p<0.05, FWE corrected). Considering extinction, no significant correlations with gray matter cerebellar volume were observed. The present findings show that reduction of storage of learned eyeblink responses with age is explained at least in part by age-dependent decline of cerebellar function. Future studies need to be performed to better understand which brain areas contribute to age-dependent reduction of extinction.

The aim of the present study was to examine somatotopy in the cerebellar cortex and a possible differential role of the cerebellar cortex and nuclei in functional outcome. Clinical findings and 3D MRI-based cerebellar lesions site were compared in a group of 90 patients with focal cerebellar lesion using International Cooperative Ataxia Rating Scale (ICARS) and voxel-based lesion-symptom mapping (VLSM). Separate analysis was performed in patients with acute and chronic ischemic lesions (n=43) and patients with acute and chronic surgical lesions (n=47). Thirty-eight patients were included after resection of a cerebellar tumor in childhood or adolescence. The most significant lesion symptom correlations were observed in the subgroup with acute ischemic lesions. Limb ataxia was significantly correlated with lesions of the interposed (NI) and part of the dentate nuclei (ND), ataxia of posture and gait with lesions of the fastigial nuclei (NF) including NI. Correlations with cortical lesions were less significant and present in the superior cerebellum only. Upper limb ataxia was correlated with lesions of vermal, paravermal and hemispheral lobules IV-V and VI, lower limb ataxia with lesions of vermal, paravermal and hemispheral lobules III and VI, dysarthria with lesions of paravermal and hemispheral lobules V and VI and ataxia of posture and gait with lesions of vermal and paravermal lobules II, III and IV. In the subgroups with chronic focal lesions, similar correlations were observed with lesions of the cerebellar nuclei, but significantly less correlations with lesions of the cerebellar cortex. Functional localization based on VLSM backs findings in previous animal and functional brain images studies in healthy human subjects. The lesion site appears to be critical for motor recovery. Lesions affecting the cerebellar nuclei are not fully compensated at any age and independent of the pathology in humans.

There is increasing evidence of a topographic organization within the human cerebellar cortex for motor and non-motor functions. Likewise, a subdivision of the dentate nucleus in a more dorsal and rostral motor domain and a more ventral and caudal non-motor domain has been proposed by Dum and Strick (2003) based on anatomical studies in monkey. In humans, however, very little is known about topographic organization within the dentate nucleus. Activation of the dentate nucleus in a verb generation task was examined in young and healthy subjects using ultra-highfield 7T functional magnetic resonance imaging (fMRI) with its increase in signal-to-noise ratio. Data of 17 subjects were included in statistical analysis. Subjects were asked to (i) read words (nouns) aloud presented on a screen, (ii) silently read the same nouns, (iii) silently generate the appropriate verbs to the same nouns and (iv) to silently repeat the names of the months. A block design was used. For image processing, a recently developed region of interest (ROI) driven normalization method of the dentate nuclei was applied. Activation related to motor speech (contrast aloud reading minus silent reading) was strongest in the rostral parts of the dentate nucleus. Dorsorostral activations were present bilaterally. Activation related to verb generation (contrast verb generation minus silent reading) was found in the ventrocaudal parts of the dentate nucleus on the right. The present findings are in good accordance with the anatomical data in monkeys and suggest that the human dentate nucleus can be subdivided into a rostral and more dorsal motor domain and a ventrocaudal non-motor domain.