Specific cerebellar regions are related to force amplitude and rate of force development

Abstract

The human cerebellum has been implicated in the control of a wide variety of motor control parameters, such as force amplitude, movement extent, and movement velocity. These parameters often covary in both movement and isometric force production tasks, so it is difficult to resolve whether specific regions of the cerebellum relate to specific parameters. In order to address this issue, the current study used two experiments and SUIT normalization to determine whether BOLD activation in the cerebellum scales with the amplitude or rate of change of isometric force production or both. In the first experiment, subjects produced isometric pinch-grip force over a range of force amplitudes without any constraints on the rate of force development. In the second experiment, subjects varied the rate of force production, but the target force amplitude remained constant. The data demonstrate that BOLD activation in separate sub-areas of cerebellar regions lobule VI and Crus I/II scales with both force amplitude and force rate. In addition, BOLD activation in cerebellar lobule V and vermis VI was specific to force amplitude, whereas BOLD activation in lobule VIIb was specific to force rate. Overall, cerebellar activity related to force amplitude was located superior and medial, whereas activity related to force rate was inferior and lateral. These findings suggest that specific circuitry in the cerebellum may be dedicated to specific motor control parameters such as force amplitude and force rate.

Figure 1

A. Recorded force trace from the force amplitude experiment showing one subject performing a single isometric contraction to the 5%, 20%, 40%, 60%, 80% MVC force levels. All contractions in the force amplitude experiment were approximately 4s long. B. A recorded force trace from the rate experiment showing one subject performing a single isometric contraction during the 0.5s Pulse, 1s, 2s, 4s, and Hold conditions. All contractions in the rate experiment had a peak force amplitude of 15% MVC.

Figure 2

A-C and D-F depict behavioral data for the force amplitude experiment and the rate experiment, respectively. A plot of group mean force amplitude (A) depicts that subjects were able to produce force close to the target level for each of the five conditions. Rate of change of force increased with force amplitude (B), while a plot of duration of force shows that the contractions were of similar length during all five conditions (C). In the rate experiment, the group mean force was similar across all conditions (D). Subjects effectively produced contractions with decreasing rate of change of force across the 0.5s, 1s, 2s, and 4s conditions (E). Contractions increased in duration across the 0.5s, 1s, 2s, and 4s task conditions (F). Error is the standard error of the mean. In some instances, the error is not apparent because the error across subjects was very small and the y-axis scale is large.

Figure 3

Voxel-wise results of the ANOVA from the force experiment (A) and the rate (B) experiment overlaid on the SUIT T1 weighted average across 20 young healthy adults. C, shows the overlap following the conjunction analysis. Blue = force experiment. Red = rate experiment. Yellow = overlap between force and rate experiment. D, shows the SUIT template for each Z slice (Diedrichsen et al., 2009).

Figure 4

Center of mass coordinates for the x (A) and z (B) dimensions are shown. The 80% MVC condition and the 4s ramp conditions were compared across left and right sides. The center of mass was quantified from each subject and condition separately using 3dclust (AFNI) in a combined ROI including cerebellar lobules V and VI and Crus I (SUIT). Error bars represent standard error.

Figure 5

Force level plotted against percent signal change in ROIs that include: left lobule V, right lobule V, left lobule VI, right lobule VI, left Crus I, right Crus I, and vermis VI. Force level is on the x-axis because this was the independent variable manipulated in the study. Each plot represents the average force/percent signal change across subjects, along with the standard error of the mean for the x and y axes. The error for force level (x axis) was small at the low force levels, and can only be visible at the high force levels. In each plot, the positive slope from the regression analysis was significant at p<.05.

Figure 6

Rate of force plotted against percent signal change in ROIs that include: left lobule VI, right lobule VI, left Crus I, and right Crus I. Rate of force is on the x-axis because this was the independent variable manipulated in the study. Each plot represents the average rate/percent signal change across subjects, along with the standard error of the mean for the x and y axes. The error for rate (x axis) was small at the low rate levels, and can only be visible at the high rate levels. In each plot, the negative slope from the regression analysis was significant at p<.05.