Abstract
Finger movements are associated with a relatively large neural representation. Passive finger movement-externally generated movement without voluntary muscle activity-is a robust approach to investigate the neural representation of kinesthesia and proprioception. While some studies have characterized the neural correlates of passive finger movement, they have relied solely on mass univariate analysis, potentially limiting sensitivity. Additionally, limited consideration has been given to stimulus duration, a factor closely tied to kinematic features such as amplitude and velocity, which more recent modeling approaches account for explicitly. Here, we reanalyzed previously published data using both univariate and multivariate analyses to examine how kinesthesia is represented in the brain in neurotypical subjects across two experiments. Systematic passive stimulation of the fingers was provided using an MR-compatible robot while functional magnetic resonance imaging data were recorded. Our analyses were conducted separately for amplitude, velocity, and direction and adjusted for stimulus duration, thereby controlling for this factor whether or not neural activation scaled with it. We provide a detailed mapping of brain areas related to these kinematic features, including sensorimotor, subcortical, and cerebellar regions. In general, multivariate pattern analysis was more sensitive than the univariate approach in identifying brain regions associated with passive finger movement. Our univariate results showed that activity in sensorimotor and subcortical areas was higher for larger amplitudes and slower velocities, which contrasted with the original study's findings, likely due to our treatment of stimulus duration as a parametric modulator. A novel result was that sensorimotor areas showed higher activation for extension compared to flexion of passive finger movement. Across kinematic features, a larger neural representation was observed for amplitude and direction than for velocity, suggesting that kinesthesia and proprioception may rely more strongly on displacement than on movement. Whereas univariate analyses are limited in addressing heterogeneity across voxels, our multivariate analyses revealed that a broader set of brain regions carried condition-related information about passive movement, based on distributed voxel activity patterns. Together, these findings extend current knowledge on how the brain represents physical kinematic features of finger movements.