Controlling a computer with thoughts?

The projects build upon ongoing research conducted in epilepsy patients who had the interfaces temporarily placed on their brains and were able to move cursors and play computer games, as well as in monkeys that through interfaces guided a robotic arm to feed themselves marshmallows and turn a doorknob.

"We are now ready to begin testing BCI technology in the patients who might benefit from it the most, namely those who have lost the ability to move their upper limbs due to a spinal cord injury," said Michael L. Boninger, M.D., director, UPMC Rehabilitation Institute, chair, Department of Physical Medicine and Rehabilitation, Pitt School of Medicine, and a senior scientist on both projects. "It's particularly exciting for us to be able to test two types of interfaces within the brain."

"By expanding our research from the laboratory to clinical settings, we hope to gain a better understanding of how to train and motivate patients who will benefit from BCI technology," said Elizabeth Tyler-Kabara, M.D., Ph.D., a UPMC neurosurgeon and assistant professor of neurological surgery and bioengineering, Pitt Schools of Medicine and Engineering, and the lead surgeon on both projects.

In one project, funded by an $800,000 grant from the National Institutes of Health, a BCI based on electrocorticography (ECoG) will be placed on the motor cortex surface of a spinal cord injury patient's brain for up to 29 days. The neural activity picked up by the BCI will be translated through a computer processor, allowing the patient to learn to control computer cursors, virtual hands, computer games and assistive devices such as a prosthetic hand or a wheelchair.

The second project, funded by the Defense Advanced Research Projects Agency (DARPA) for up to $6 million over three years, is part of a program led by the Johns Hopkins University Applied Physics Laboratory (APL), Laurel, Md. It will further develop technology tested in monkeys by Andrew Schwartz, Ph.D., professor of neurobiology, Pitt School of Medicine, and also a senior investigator on both projects.

It uses an interface that is a tiny, 10-by-10 array of electrodes that is implanted on the surface of the brain to read activity from individual neurons. Those signals will be processed and relayed to maneuver a sophisticated prosthetic arm.

"Our animal studies have shown that we can interpret the messages the brain sends to make a simple robotic arm reach for an object and turn a mechanical wrist," Dr. Schwartz said. "The next step is to see not only if we can make these techniques work for people, but also if we can make the movements more complex."

In the study, which is expected to begin by late 2011, participants will get two separate electrodes. In future research efforts, the technology may be enhanced with an innovative telemetry system that would allow wireless control of a prosthetic arm, as well as a sensory component.

"Our ultimate aim is to develop technologies that can give patients with physical disabilities control of assistive devices that will help restore their independence," Dr. Boninger said.