Motor planning brings human primary somatosensory cortex into action-specific preparatory states

Motor planning plays a critical role in producing fast and accurate movement. Yet, the neural processes that occur in human primary motor and somatosensory cortex during planning, and how they relate to those during movement execution, remain poorly understood. Here we used 7T functional magnetic resonance imaging (fMRI) and a delayed movement paradigm to study single finger movement planning and execution. The inclusion of no-go trials and variable delays allowed us to separate what are typically overlapping planning and execution brain responses. Although our univariate results show widespread deactivation during finger planning, multivariate pattern analysis revealed finger-specific activity patterns in contralateral primary somatosensory cortex (S1), which predicted the planned finger action. Surprisingly, these activity patterns were as informative as those found in contralateral primary motor cortex (M1). Control analyses ruled out the possibility that the detected information was an artifact of subthreshold movements during the preparatory delay. Furthermore, we observed that finger-specific activity patterns during planning were highly correlated to those during execution. These findings reveal that motor planning activates the specific S1 and M1 circuits that are engaged during the execution of a finger press, while activity in both regions is overall suppressed. We propose that preparatory states in S1 may improve movement control through changes in sensory processing or via direct influence of spinal motor neurons. Significance statementMotor planning is important for good behavioral performance, yet it is unclear which neural processes underlie the preparation of the nervous system for an upcoming movement. Using high-resolution functional neuroimaging, we investigated how motor planning for finger presses changes the activity state in primary motor and primary somatosensory cortex, and how brain responses during planning and execution relate to each other. We show that planning leads to finger-specific activation in both M1 and S1, which is highly similar to the finger-specific activity patterns elicited during execution. Our findings suggest that S1 is being specifically prepared for an upcoming action, either to actively contribute to the outflowing motor command or to enable action-specific sensory gating.