MOTORLAB
University
of Pittsburgh
Image Caption Abstraction of brain network activity during learning. Each wave represents a single neuron and its orientation and height reflect its contribution to a brain-controlled movement in virtual reality. The tip of each wave represents the learned effect of this contribution in response to an imposed perturbation to the algorithm used to interpret the recorded brain signals. Even though individual cells have different preferred directions (as represented by their position relative to an origin located over the horizon), their directional responses to the perturbation are generally consistent.

As a systems neurophysiology lab, we are interested in the way neural activity drives behavior. Our goal is to describe organizational principles of the time-varying relation between the firing rates of neurons and the behavior this activity generates.

Specifically, our research program centers on the relationship between cerebral cortical activity and arm movement. Over the last 25 years, we have found that there is a very good representation of the arm’s trajectory in the collective firing pattern of frontal cortical activity. This makes it possible to predict the detailed time course of arm, wrist and finger movement that contains many of the behavioral invariants that are characteristic of movement.

Research projects in our laboratory are based on this dynamic representation of behavior. The parameter of time is fundamental in our consideration of movement. We have shown that movement generation takes place continuously—the cortical prediction of trajectory precedes movement execution with a time interval that is dependent on the figural components of the hands’ path.

Laboratory studies range from muscle activation and limb mechanics, learning mechanisms, and information metrics to object-hand interaction. Results from this basic science research are used to design and implement upper-extremity neural prosthetics. Our prosthetic work began in the early ‘90s and has progressed from demonstrations in monkeys to implementation in a paralyzed human subject. In 2012, Jan Scheuermann used neural signals tapped byelectrode arrays implanted on the surface of her brain to operate a high- performance robotic arm and hand. She was able to move the prosthetic arm, wrist and fingers to perform tasks of daily living, showing that this technology will be able to restore useful function to those who cannot move their arms and hands.