Mechanisms of human prehension

Two neural network models of human prehension are reviewed. The first model simulates coordination of the reach and grasp components of prehension. Simulations of this model produce realistic timing of the velocity peaks of reach and grasp; realistic scaling of peak grasp size with reach velocity and target grasp size; and realistic adaptation of reach and grasp to perturbations of initial grasp size, target grasp size, and target reach position. The second model simulates control of the load and grip forces of prehension. Simulations of this model produce realistic grip force anticipation; realistic learning of adequate grip forces for different object weights and textures; and realistic timing of peak grip forces. Future studies will investigate learning of adequate peak grasp sizes and timing, and reactions of grip forces to unexpected changes of object weight and texture