Our goal is to understand the basic neural and perceptual mechanisms underlying the human and simian ability to recognize and manipulate objects with the sense of touch. To correlate neurophysiological activity with hand movements, we use state-of-the-art computer-data-acquisition techniques, including digital video. Current projects include: 1) coding of motion and its direction across the skin; 2) neural representation of dot arrays forming textured surfaces; 3) cortical analysis of the size and shape of hand-grasped objects; and 4) spatial organization of receptive fields of cortical neurons receiving inputs from different classes of cutaneous mechanoreceptors.

We found that cortical processing involves feature extraction through convergent central organization of functional neuronal assemblies. These modules represent specific skin regions, such as individual fingers, the tips of several adjacent fingers, or fingers and palmar skin. However, individual neurons within a module respond to particular features, such as direction of motion or posture of hand. Neuronal-response quantification may reveal how parallel processing of sensory information in adjacent cortical modules helps distinguish object size, shape, and texture. Elucidation of such neural networks in the brain provides a basic architecture for object recognition in the sense of touch. Such findings are useful for developing intelligent hands with tactile sensors for prosthetic devices or robotics. They also have important clinical applications, for instance, creating quantitative sensory-function tests in patients with neurological disorders or peripheral nerve injuries, or in sensory substitution aids for visually and/or hearing-impaired individuals.