Properties of macaque ventral premotor cortex during grasping

Abstract

The hand is one of the most sophisticated biological motor systems and understanding the control strategies used by the brain to move this complex apparatus represents a major challenge. Previous studies have given evidence of multiple cortical representations of hand movements including primary motor cortex, supplementary motor area, inferior area 6 and parietal cortex. These findings raise questions about the specificity of each of theses areas for the planning and control of distal movements. In the present study, the main focus was to investigate the properties of area F5 (rostral part of inferior area 6). Firstly, we investigated the specificity of the response of F5 neurons to two parameters, grip type and target orientation, during a delayed grasping task. This task was divided into well defined periods that allow the analysis of the neural response during different phases of the action, namely the object observation, the planning of the movement and the movement execution. Secondly, we investigated the representation of partial instruction information by separating the instruction of the orientation from the instruction of the grip type in distinct task epochs. Thirdly, a decoding simulation was performed in order to attest the possibility of decoding grasp movement intentions from area F5 for the possible application of a prosthetic hand control from brain signals. Specifically, we recorded neural activity in two macaque monkeys while they were presented with a handle that could be rotated in five possible orientations (upright, 25 and 50 degrees to the right or left). Simultaneously with the object presentation, a colored light instructed how the handle had to be grasped, either with a power or with a precision grip. Our results revealed that the grip type and the object orientation are both encoded in area F5. Their representation was similar during the instruction, but while the representation of object orientation was maintained constant, the representation of grip type significantly increased during the movement execution. These results suggest a major role of area F5 for shaping the fingers during grasping movement while its role for the positioning of the hand in the correct orientation might be reduced. Furthermore, cells with different tuning onset for grip type and orientation had been found. These different types of cells also showed differences in the simultaneous encoding of grip type and orientation. Moreover, the cue separation task revealed that the orientation representation was present in F5 even without a grip type instruction, but that the grip type was not encoded in F5 in the absence of the presentation of an object. Finally, the decoding simulation using a Bayesian classifier showed that grip type and orientation could both be decoded from area F5, but the performance was better for decoding the grip type than the orientation. In sum, the present thesis brings new insight to the representation of hand movement in area F5, in particular for the combined encoding of object orientation and grip type. It also reveals that hand movements can be decoded from higher order planning areas and that area F5 could be suitable for the implementation of a brain-machine interface for hand grasping which might have potential value for future applications in paralyzed patients.