Experiments On The Ethological Organization Of Motor Cortex

Abstract

This thesis explored several aspects of the cortical organization and representation of movement in monkey motor cortex. The experiments were based on the finding that long-duration stimulation applied to the monkey motor cortex evokes coordinated movements that can resemble actions from the monkey's behavioral repertoire (Graziano, Taylor, and Moore 2002). The thesis presents three experiments.

First, we used electrical stimulation of the monkey motor cortex to more carefully quantify the properties of stimulation-evoked movements. We found that arm movements evoked by stimulation share many properties with natural voluntary movements including bell-shaped velocity profiles, proper speed-distance scaling, inter- joint coordination to stabilize the position of the hand in space, and some ability to compensate for the effects of a weight attached to the hand. Movements evoked by stimulation were topographically organized on the cortical surface by the set of body parts recruited, the final position of the hand in space, and the type of behavioral action evoked.

Second, we used single neuron recording to study the tuning of single neurons in primary motor cortex during naturalistic arm movements. We found that the behavior of single neurons over the entire naturalistic dataset was best explained by tuning to a preferred posture of the arm. On average, the preferred posture of single units and the posture evoked by electrical stimulation at the same cortical site were in rough agreement. The neurons were also tuned to other aspects of movement. One possibility is that control signals represented by motor cortex neurons are multiplexed and tuned to the varied features that characterize statistical biases in the monkey's natural motor behavior.

Third, we used computational modeling methods to study the principles of topographic organization in the motor cortex. We found that the overall topographic organization of motor cortex was largely explained by a model that mapped a high- dimensional representation of the behavioral repertoire onto the two dimensional cortical surface. We were thus able to provide a parsimonious account for many otherwise idiosyncratic features of cortical organization.