Distinctions in prosthetic control

Written byEdyta Zielinska

| 2 min read

The popular press was a-buzz this week with reports of a technique that could allow an amputee to move her prosthetic arm with linkurl:her mind.;http://www.nytimes.com/2009/02/11/health/research/11arm.html?_r=1&scp=1&sq=todd%20kuiken&st=cse But in fact, the technology, developed by Todd Kuiken and his group at Rehabilitation Institute of Chicago, doesn't attempt to read a patient's thoughts -- at least not directly. The type of control Kuiken's group is working on is one step removed. In a woman who had lost her left arm in a car accident and four other amputees, they took the nerves that used to control elbow, wrist and finger movements and surgically threaded them into muscles in the chest. Once the patient healed, the scientist placed electrodes on the area of the chest above those muscles, and recorded the electrical signals through the skin. The signals from those nerve endings were then routed to a computer that interprets them, using them to control the motion of a prosthetic arm.

Researchers elsewhere are indeed trying record signals directly from brain tissue in patients with severe paraplegia. These patients have recording electrodes implanted in their brains that read signals that correspond to their intentions to move. (You can read more about that technology in our linkurl:January issue.;http://www.the-scientist.com/2009/01/1/32/1/ ) Kuiken demonstrated his nerve implantation approach several years ago, in a patient named Jesse Sullivan who had lost both of his arms. The advance on the technique linkurl:published;http://jama.ama-assn.org/cgi/content/short/301/6/619 this week in the Journal of the American Medical Association provides much finer motion control in prosthetic arms, though it needs a better prosthetic arm than those widely available to work. "The difference is not the surgery," Kuiken told me. Rather, Kuiken's group tinkered with the computer that translates a patient's intention into action in the robotic arm. Originally, the group had simply recorded the amplitude of the signal that came from the patient's residual nerve. Now, however, they were recording many other features of the electrical signal, and interpreting a pattern of information rather than a single blip. How it works: the patient, with electrodes stuck on the surface of her skin, is asked try to move her wrist. The brain activation associated with this intention to move is sent down to the residual nerve that would normally have controlled her wrist, but which has now been transplanted to her chest. The electrodes on her chest pick up the signal and send it to the computer, where the pattern is analyzed and labeled as a "wrist moving pattern." The computer essentially learns the electrical shape of the patient's intention. The same process is repeated for other complex motions. Once the patient has trained the computer, she's ready to use those signals, created just for her, to control an actual prosthetic. To move her arm in a particular direction, she mentally makes the motion. That brain signal is then analyzed for its signature pattern by the computer and routed to the electronics of the prosthetic arm to make the motion. Because prosthetic arms available today don't have the capability to receive these more complex commands, Kuiken's team used two new robotic arms, designed as part of the Defense Advanced Research Project Agency's Revolutionizing Prosthetics program. "We got to take them both for a ride," said Kuiken. The new prosthetic arms -- which aren't yet available for commercial use -- have motorized shoulder, elbow and wrist movement, allowing for a wider range of movements.
**__Related stories:__***linkurl:Of cells and wires;http://www.the-scientist.com/2009/01/1/32/1/
[January 2009]*linkurl:Re-engineering humans;http://www.the-scientist.com/2007/3/1/28/1/
[March 2007]