13.7 News & Views MH

Using our thoughts to control a computer or robot used to be the realm of science-fiction writers. But scientists have been making concerted efforts to develop the technology required to convert brain signals into commands, to support communication, mobility and independence for paralysed people. In this issue, two papers shift the notion of such ‘implantable neuromotor prosthetics’ from science fiction towards reality. Hochberg et al. (page 164) describe the first implantation of electrode arrays into the brain of a paralysed man; these allowed him to use his thoughts (motor intentions) to directly control devices such as a computer mouse. In addition, Santhanam and colleagues (page 195) describe a new software approach for extracting intended actions from the neural activity in the brain of monkeys that dramatically improves the potential speed and performance of implantable neuromotor prosthetics. Although it may someday be possible to reconnect damaged neural pathways by directing the regrowth of neurons, neuroprosthetics provide another potential approach to permit individuals with severe neurological injuries to interact with the environment. Damage to the nervous system means these individuals lose their motor control — that is, they can no longer directly control their muscles. Neuroprosthetics aim to bypass this damage by recording brain activity that reflects the individual’s motor intentions. These signals are then used either to reanimate paralysed muscles using electrical stimulation, or to control physical devices directly, such as artificial limbs, computer cursors or wheelchairs. Several different approaches have been developed ranging from non-implantable technologies that record electroencephalographic (EEG) activity using removable electrodes placed on the scalp surface, to implantable devices that use microelectrodes to detect the activities of individual neurons. Implantable neuromotor prosthetics build on basic research looking at the neural basis of the planning and control of movement, predominantly in monkeys. A brain region of particular interest for neuroprosthetic research is the primary motor cortex, a major region involved in controlling voluntary movements. Others include the premotor and posterior parietal cortex, which are principally involved in planning movements, providing instructions for the motor cortex to then act on. In monkeys, the activities of large numbers of neurons can be recorded simultaneously and be used to predict motor intention and limb motion, or to move cursors on a computer screen. Animal experimentation remains essential for testing and developing implantable neuroprosthetics, but obviously at some point the great leap into human patients needs to be made. To this end, Hochberg et al. recruited a man who can no longer move his limbs because his spinal cord is completely severed. They implanted an array of tiny electrodes into his primary motor cortex, and tested whether the activity of neurons recorded there could control prosthetic devices. Remarkably, the implanted electrodes allowed the patient to control a computer cursor and rudimentary movement of robotic devices (the associated videos are online in Supplementary Information). This is not the first neuromotor prosthetic that has been implanted into a person — a previous experiment used a couple of implanted electrodes to generate limited horizontal control of a cursor. But this study reports several significant advances. First, even though the patient had been paralysed three years earlier, the neural activity in his primary motor cortex seemed relatively 13.7 News & Views MH 7/7/06 5:40 PM Page 141