Functional MRI of the deep cerebellar nuclei

Background: Dum and Strick (J Neurophysiol; 2003; 89: 634–639) proposed a division of the cerebellar dentate nucleus into a „motor“ and „nonmotor“ area based on animal data. We asked the question if motor and nonmotor domains of the dentate nucleus can be found in humans using fMRI. Dentate activation was compared in motor tasks of different complexity and two cognitive tests of executive function, that is verbal working memory and planning.

Methods: A total of 23 right-handed healthy subjects (7 male/16 female, mean age of 27,8±6,3 years) participated. Only data of individual subjects were included in final analysis which showed significant activation of the cerebellar nuclei in at least one of the tasks. A Siemens 1.5 T Sonata system and an 8 channel head coil were used for functional scans. Preprocessing and statistical analysis of the imaging data was performed with statistical parametric mapping software, version SPM5, implemented in MATLAB.

A block design was used for the experimental tasks. All tasks were performed with the right hand. Each of the four conditions was compared with an adequate control condition.

• Simple finger tapping: Subjects used their right index finger and pressed one key of an optical response keypad at a frequency of 3Hz.

• Complex finger tapping: The four keys of the keypad had to be pressed in the following sequence: [1–2-3–4-4–3-2–1] at a frequency of 3Hz.

• Verbal working memory: Subjects had to memorize a four number sequence, consisting of the numbers 1 to 4 (e.g. 2–3-2–4), and gave the answers using the keypad.

• Planning task: Subjects navigated through a labyrinth with a custom made optical response joystick.

Results: Significant dentate activation was found in twelve subjects which entered group analyses. Ventrocaudal activation of the ipsilateral dentate nucleus was observed in all experimental conditions (for example Fig.1: complex finger tapping). In the cognitive tasks additional dentate activation of the contralateral side was found (for example Fig.2: verbal working memory).

Discussion: Dentate activation for motor and cognitive tasks was predominantly found in the ventrocaudal part of the dentate nucleus corresponding to the nonmotor area proposed by Dum and Strick (2003). Methodological limitations using 1.5 T MRT may be one possible reason for the lack of activation of the dorsal portion (that is the proposed motor area) of the dentate nucleus.

Supported by DFG TI 239/9–1