The ultimate objective of this project is to assist the rehabilitation of sensory function in the quadriplegic hand and arm. In particular, this project will provide information necessary for implementing artificial kinesthetic sensory feedback in a hand-grasp neuroprosthesis. In fact, the results of the proposed experiments will provide such information at three levels. First, the data will constitute an extensive, unified, empirical description of kinesthetic perception. Previous investigations have shown that the perceived force, effort and displacement from a voluntary contraction interact to form a composite sensation. Although the results from different studies agree qualitatively, they cannot be compared quantitatively due to differences in instructions, paradigms, apparatus, subjects and test locations. As a result, the apparent contributions of the component sensations to the total sensation vary considerably. In order to obtain a coherent and mutually-reinforcing set of observations, many of these earlier experiments will be replicated under more uniform conditions. For example, most of the experiments will be performed using the same task (three-finger pinch, elbow flexion or both), apparatus and subjects, when possible. The experiments will include: direct estimation and bilateral matching of perceived force, effort, and displacement as a function of nominal strength, muscle length, atrophy and fatigue; threshold discrimination and matching of displacement, joint angle and orientation; and bilateral matching of displacement, force and effort as a function of variations in the load. Second, beyond providing an empirical description of kinesthetic perception, some of these experiments will address fundamental questions regarding the internal representation of sensations. In part, the experiments will test the theory of absolute sensation magnitude, which has successfully accounted for some limited auditory loudness results, but has not been tested widely and has not been applied to kinesthesis. In addition, the experiments will determine the appropriate representations for force, effort, and three variables related to kinesthetic """"""""space"""""""": absolute joint angle, angular displacement about a joint, and the orientation of a limb with respect to a fixed coordinate frame. Third, and most important, the experiments have been designed to test a quantitative, psychophysical model of kinesthetic perception. In this model, force, displacement and the efferent command are transformed into sensations via power functions, and the net kinesthetic sensation is the weighted sum of the individual components. The model is expected to provide a framework or """"""""blueprint"""""""" for the development of artificial kinesthetic sensory feedback. However, though data from previous experiments provide a preliminary empirical validation of the model, the rigorous series of experiments contained in this project must be completed before the model can be accepted with confidence.
