Paul Cisek - US grants

The purpose of the proposed project is to evaluate a model of cortical motor control through a series of neurophysiological experiments. The model tested has been developed based on a diverse set of physiological, anatomical, and psychophysical data, with particular focus on data about cell responses in cortical areas 4 and 5 during voluntary and passive movements. Because the model proposes specific functional roles to different neural regions, it makes specific predictions regarding cell responses in these regions during novel experimental paradigms. The particular hypotheses to be tested are the proposals that tonic activities in area 4 correspond to a descending command that combines both position and force information, and that tonic activities in anterior area 5 correspond to a representation of current limb position. These hypotheses make predictions regarding the difference in cell responses between areas 4 and 5 during tasks which dissociate force-and-position information from pure position information. Two experimental paradigms are proposed to test this, one involving a comparison between isometric movements and movements with a pendulum, the other involving movements in the presence of obstacles. Aside from its relevance to the model being tested, the data obtained is expected to contribute to the understanding of cortical function during voluntary movement.

The purpose of the proposed project is to investigate the cerebral cortical mechanisms involved in the specification and selection of multiple potential motor actions. The motivation for this work lies in a broad theoretical framework based on parallel neural mechanisms of "action specification" (which specify the parameters of potential actions) and mechanisms of "action selection" (which choose between several possible courses of action). In this framework, neural processing of sensory input can lead to the simultaneous specification of several potential motor actions, one of which may be released into execution (selected) depending on the behavioral context. In our experiment, we train a monkey to make reaching movements to spatial targets after an instructed delay period during which two potential targets are presented. During the first half of the instructed delay, the monkey does not know which target will be the correct one. During the second half, a non-spatial cue indicates the correct target. The question is: In how much detail does the monkey prepare the two potential movements while he still does not know which will be the correct one? That is, does he merely remember sensory representations of the targets or does he prepare (specify) both potential movements with the same degree of detail with which a single unambiguous movement can be prepared? Preliminary results favor the latter possibility, supporting the framework of parallel action specification and selection.