Neural Prosthetics

The capture and interpretation of brain signals from monkeys trained to think about movements and refrain from actually making these movements could lead to better communications and control devices for people with paralysis, according to a new study in the 9 July 2004 issue of the journal Science.

The researchers recorded signals from brain neurons in the monkeys' parietal cortex and premotor cortex that encode the goals of reaching movements. They then decoded these signals and used the information to position a cursor on a computer screen.

Past studies have demonstrated that information from neurons coding physical movement instructions can be used to control prosthetic devices. The new research, however, looks at prosthetic control using higher-level brain signals. When the monkey's abstract thought was consistent with the requested movement — and the monkey did not move during the thought — the monkey received a reward. Performance of the task improved over a period of weeks.

In addition to collecting information on the goal, the researchers simultaneously recorded the reward preferences of the monkeys. When the monkeys knew that accurately thinking about the requested movement would yield a preferred reward, their cursor-positioning thoughts were more accurate. Similar goal and preference information from patients with paralysis could be useful, according to author Richard Andersen. He proposed that future technologies based on this type of research could allow a paralyzed person's goal-directed thoughts, as well as preferences, to be actualized with smart machines.

8 July 2004