MIT team finds that the ratio of component atoms is vital to performance.
Over the last month Kelli Keefe has been testing a system designed and developed by MIT engineers that allows Ms. Keefe, who is paralyzed from the waist down, to stand for virtually unlimited amounts of time and walk short distances (currently with the aid of parallel bars).
Although clinical tests have only just begun, Professor William K. Durfee of mechanical engineering and the seven MIT students he is working with are hopeful that one day the system will allow people who are paralyzed to stand while delivering a paper, walk to a colleague's office, and more.
The system combines electrical stimulation of the muscles with a brace around the legs, hips, and waist. It is unique because the brace is fitted with small computer-controlled "brakes" that work much like the hand brakes on a bicycle.
"We use Kelli's electrically stimulated muscles to provide the power, and the brakes to control the resulting movement," said Dr. Durfee, who is the Brit and Alex d'Arbeloff Associate Professor of Engineering Design.
Professor Durfee and his team have been working on the system for about two years. The graduate students involved are Michael Goldfarb, Jeffrey T. Chiou and Heather L. Beck (all in mechanical engineering). Undergraduates are Aaron M. Barzilai (a senior in mechanical engineering), Angela Hsieh (a junior in electrical engineering and computer science), and sophomores Carrie E. Perlman and Peter W. Kassakian of mechanical engineering. The work is funded by the Department of Veterans Affairs.
RESTORING PHYSICAL FUNCTION
About 150,000 people in the United States are paralyzed from the waist down (paraplegia) or neck down (quadriplegia), primarily because of accidents in sports and with cars and guns.
And every year about 8,000 more people are paralyzed-people with a mean age of 23. "So we're talking about people who are very young, with a lifetime ahead of them, yet they're in a wheelchair," Professor Durfee said.
As a result, for many years scientists have been working to restore physical function to these people via electrical stimulation of the muscles. (The specific technique is known as functional electrical stimulation, or FES.)
In people with paraplegia the pathway between the brain and the spinal cord is cut, so the brain can't send a signal to, say, the leg muscles, telling them to contract. With electrical stimulation "we effectively replace the signal coming down with an artificial one," Professor Durfee said.
Generally, in FES electrodes are placed over the muscles. (Sometimes they are implanted in them, but according to Professor Durfee this technology is still at an early stage). A computer then coordinates where and when electrical current is applied through the electrodes so that the muscles contract and relax in the proper order-and the person walks.
COMPLEX SET OF PROBLEMS
"On the surface [this technique] sounds fairly simple, but the more you dig into it the more complex it becomes," Professor Durfee said.
For example, the number of muscles scientists can access with FES is much lower than the total number available. (One specific example: the muscle that makes the hip come forward is deep inside the hip, so it can't be stimulated with external electrodes.)
Further, to really control the system the engineers want to make a variety of physical measurements, like the angles in a joint, that they can then put into the computer. "The more measurements we can make, the easier it is to control the system," Professor Durfee said.
But such measurements are often difficult-or impossible-to make. For example, said Professor Durfee, "we can't measure force inside a muscle, so we only have a vague idea of how much we're getting out of a muscle."
The most serious drawback of FES, however, is that muscles stimulated artificially tire very rapidly.
The MIT engineers are taking two approaches toward solving these problems. First, they have developed and are refining their hybrid FES system with its brace and computer-controlled brakes. Second, they are doing basic research to create mathematical models of how muscles behave when stimulated.
A BRACE WITH BRAKES
In the first approach, Professor Durfee said, "we're applying our expertise as mechanical designers to the whole problem."
For example, the researchers decided early on that it would be easier to control a mechanical device-the brace via its "smart" brakes-with properties that are well known, than to control the muscles, with properties that are often vaguely known.
Explained Professor Durfee: "If we wanted to control the speed with which a knee extends using FES alone, we would have to adjust the stimulation [of the muscles] just right. And that's hard to do.
"Now we're saying, `let's not control the muscles-let's use them as on/off sources of power [via FES], and use the brakes to control movement.' "
As a result, the researchers are hopeful that the hybrid system "will lead to gaits that are more coordinated and faster," Professor Durfee said.
The system also addresses the muscle-fatigue problem associated with FES.
When a paralyzed person using FES stands, the muscles are being used. And since those muscles tire rapidly, that person can only stand for a short period of time.
With the MIT system, however, "once the person is standing we can lock the brace and turn off the muscles, so he or she can stand for virtually unlimited periods of time," Professor Durfee said. Furthermore, "now we only need to turn on the muscles when the person is taking a step, and that cuts muscle use way, way down."
The brace was developed in large part by graduate student Michael Goldfarb. "He's a superb mechanical designer," Professor Durfee said.
Basic research on how muscles behave when stimulated is the second major part of the MIT program. Specifically graduate students Jeffrey Chiou and Heather Beck have developed and are continuing to develop mathematical models that describe these movements. Such models are incorporated into the computer control system that coordinates the FES/brace system. The better the models, the better the control and the smoother the gait.
Eventually, the researchers hope to use the basic computer control system they develop as a template to help physical therapists design customized versions. (Each person fitted for the FES/brace system will require a different computer control system based on variables like weight, leg length, and the strength of muscle contractions).
Professor Durfee emphasized that an important part of the team's basic research is verifying the mathematical models in the lab. "You have to check them with reality," he said. As a result, he continued, "our whole group is used to putting electrodes on to be subjects for stimulation experiments." (The sensation feels "a little like if you've accidentally put your finger in an outlet and don't get a shock, but kind of a buzzing.")
Of course, tests must also be made with people who are disabled. Over the last two years people with paraplegia have volunteered to help make a variety of measurements at the West Roxbury VA Medical Center and at MIT's Eric P. and Evelyn E. Newman Laboratory for Biomechanics and Human Rehabilitation.
Over the last month, Ms. Keefe has been testing the FES/brace system at the VA (three other people with paraplegia will also be testing the system). At the VA, the MIT team has been working closely with biomedical engineer Allen Wiegner (who has an MIT PhD in electrical engineering) and physical therapist Nancy Walsh.
Although it's too early to report any quantitative results, "after the first few tests the system appears to be doing a great deal of what we wanted it to," Professor Durfee said, "particularly the ability to lock up the brace while Kelli is standing."
Further, he said, "on a subjective level Kelli says that the brace gives her more confidence."
What's ahead? "Lots and lots of refinement," Professor Durfee said.
Although the MIT group hopes that one day their hybrid system could be useful to a number of people, Professor Durfee notes that the system has some drawbacks.
For one, right now the brace is clunky, difficult to put on, and is tethered to a computer via many wires. Eventually the researchers plan to shift their efforts to redesigning the system for the consumer, but "first we want to show that it works," Professor Durfee said.
Further, the MIT system and any other using FES will always require the user to use a support such as a cane or walker, because the system will never be able to compensate for every human movement that could throw the system off balance.
Professor Durfee also noted that the system will not be inexpensive, so its general acceptance will depend on whether insurance companies will pay for it. That, in turn, will depend on the system's benefits for people with paraplegia. "And it's still too early to tell how many people the system will help and the kind of function it will provide," Professor Durfee said.
Overall, he said, "it's important to have a healthy degree of skepticism about the ultimate goals of the technology and what's achievable."
Regardless of these drawbacks, the MIT team is optimistic about the future of their system.
Said Professor Durfee: "It's very exciting to go down to the VA and watch Kelli, who's been in a wheelchair, work her way down the parallel bars. And I think she's excited, too."
A version of this article appeared in the May 12, 1993 issue of MIT Tech Talk (Volume 37, Number 32).