Brain control of sensorimotor prostheses

The introduction to this volume outlines the possibilities inherent in utilizing electronic interfaces with the brain to alleviate problems of paralysis, such as that caused by spinal cord injury. The possibility of using electroencephalographic recordings for this purpose has been explored, but the amount and specificity of the information that can be extracted in this way is questionable. As an alternative, it would theoretically be more efficient, and more elegant, to utilize control information extracted from the part of the brain that normally is directly involved in processing “commands” for voluntary movement of the arm. In fact, the information contained in the motor areas of the brain is (by definition) completely sufficient to reproduce all of the normal movements of the body. It has long been considered that paralysis victims might benefit from such an ability to control devices directly from the brain. It would of course be preferable to utilize non-invasive techniques for recording the brain activity necessary to control these devices, but no currently available technique can provide sufficient information for control of external movement in real time. Electroencephalography (EEG), for example, involves recordings from the scalp, and thus can only detect the common electrical dipoles emerging from billions of individual neurons in wide areas of the subjacent brain. Though it has commonly been used for measuring global brain rhythms, it has not been particularly useful for extracting brain information specific to a particular limb movement. As such, when EEG is used as a brain-computer interface, the subjects must normally learn to control the expression of these global rhythms. Such approaches have been successfully utilized for tasks that are not time-critical, such as selecting letters by using EEG recordings to move a cursor across a computer screen.

Duke Scholars

Author Miguel Angelo L. Nicolelis Neurobiology