Neural Prosthetic Systems

The brain receives input from all the senses and can transmit signals to the spinal cord for the execution of body movements. However, injury or disease can deprive the brain of sensory input or can interrupt output pathways, resulting in motor paralysis. The field of neural prosthetics aims to restore motor or sensory function through output and input pathways other than the brain's normal channels.

Brain stem strokes and diseases, such as amyotrophic lateral sclerosis (Lou Gehrig's disease), can lead to the most severe form of paralysis, as patients can lose all movement ability and become “locked in.” Cervical spinal cord injuries (at the neck) also disrupt communication channels between the brain and spinal cord and can lead to quadriplegia (paralysis of all four limbs). All these patients have functioning brains and can still think about making movements and form the associated motor intentions, but they cannot execute these intentions. Motor prosthetic systems (MPSs) aim to harness these intentions and thoughts by recording their neural representation directly from the brain and relaying them to algorithms trained to control external devices, such as robotic arms or computer software. Ideally, in patients with spinal cord injuries, these devices may restore movements of the patients' own limbs by reconnecting the brain to the body below the injury, acting as a bridge that bypasses the severed section of the spinal cord.

An MPS is a framework that encompasses devices or programs to (a) record neural signals, (b) interpret the neural signals, and (c) control external devices. Thus, this field is highly interdisciplinary, requiring research across many branches of medicine and engineering, including neurosurgery, neuro science, computer science, robotics, materials science, and microfabrication. This entry focuses on signals recorded from inside the brain from areas related to movement planning and execution. Three areas in particular have been shown to contain useful signals for MPS: the motor cortex (M1), the parietal cortex, and the dorsal premotor cortex (dPMC). These regions form a network for visually guided reaching by transforming visual input about the location of the target into commands that move the limb toward the target. Studies have shown that this network is affected by the reward associated with a target and can participate in decision making. The viability of this network for prosthetic control is enhanced when one considers that these areas are also engaged by cognitive processes, such as movement imagination. This entry first describes the neural signals in these areas and discusses some of the studies that utilized these signals for prosthetic control. It also describes the clinical trials currently under way. Then, this entry considers some of the challenges that have to be overcome and incorporated into the future development of flexible and effective MPSs.