Professor, Bioengineering University of Pittsburgh Phone: (412) 383-7021 Email: abs21@pitt.edu Individual Website: http://motorlab.neurobio.pitt.edu/

Ph.D., University of Minnesota

Research Interests

I am interested in the brain processes that subserve volitional movement. My research program is divided into two areas; basic cortical mechanisms subserving arm movement and applied bioengineering of cortical output. The results of our basic research show that a large population of cells in frontal cortex is active during every arm movement. This population codes for details of the arm’s trajectory (path of the hand through space and time). The code is instantaneously predictive and the interval between the population code and movement trajectory is about 100 ms. To explore this continuous signal more completely, we have been looking at drawing movements which generate large and continuous changes in the arm’s direction and speed. Again, the population code predicts the arm’s trajectory. The prediction is accurate enough to recover the shape of the drawn object as well as many of the psychophysical characteristics of drawing. We have constructed an accurate biomechanical model of the arm, allowing us to look at parameters such as torque, muscle-induced acceleration and joint-angular velocity at the seven proximal degrees-of-freedom in the arm. Interestingly, we have found that joint angular velocity during 3D-reaching is highly correlated to the velocity of the hand, a finding that might not be expected because of the excess degrees-of freedom in the arm. Single-cell activity was almost equally predictive of joint-angle velocity as it was of the hand’s velocity.

By extracting accurate trajectories from areas of the frontal cortex, we have probed perceptual processes. Using a virtual reality environment, we trained monkeys to draw in free space using a cursor to represent hand position. As they draw ovals we change the hand-cursor gain in one dimension, forcing them to move their arm in a circle to draw the visualized oval. Human subjects do not perceive the gain change and believe their arm is moving in an oval. Examining the neural-derived trajectories, we found that the extracted information from primary motor and dorsal premotor cortex followed the arm path, while that from ventral premotor cortex followed the visualized (perceived trajectory).

These brain-derived trajectory signals have also allowed us to develop a practical prosthetic application. Using the population signal composed of multiple, individual unitary activity recorded in real-time, monkeys can move objects in a virtual environment. These animals are able to move a ball in a 3D display rapidly and accurately to novel targets, using the extracted signal in the absence of arm movement or muscle contraction. This is the first step in an on-going project to use signals derived from electrodes chronically implanted in immobilized subjects to control mechanical devices such as robot arms or wheelchairs, or to activate paralyzed muscles with functional electrical stimulation. These results show that subjects rapidly change their pre-existing patterns of neural activity with biofeedback to generate more accurate control signals. This is an exciting development in our process of uncovering novel concepts of cortical function during volitional movement.

Recent Publications

• Schwartz AB, Taylor DM, Helms Tillery, SI: Extraction algorithms for cortical control of arm prosthetics. Curr Opin Neurobiol 11: 701-707: 2001.
• Reina GA, Moran DW, Schwartz AB: On the relation between joint angular velocity and motor cortical discharge during reaching. J Neurophysiol 85: 2576-2589, 2001.
• Schwartz AB Moran DW: Arm trajectory and representation of movement processing in motor cortical activity. Eur J Neurosci 12: 1851-1856, 2000.
• Schwartz AB: Direct cortical representation of drawing movements. Science 265: 540-543, 1994.