Kevin Englehart, a researcher at the University of New Brunswick's Institute of Biomedical Engineering, works on an artificial hand.Brian Atkinson
The University of New Brunswick's Kevin Englehart spent four years on the most ambitious bionic arm project in the world. The goal: a prosthetic device amputees could control with their thoughts.
When the project formally wrapped up in 2009, he and his colleagues in the United States had made progress toward a fully bionic arm. He is now applying those lessons to a more modest, yet perhaps more practical endeavour, an artificial hand that will be more sophisticated and functional than anything currently available.
Amputees will be able to control it using the muscles of their lower arm. The idea is that they will be able to pick up a cup of tea or turn off a light switch by contracting the same muscles they did before they lost their hand; the subtle movement of those muscles will direct the prosthesis.
"We have developed a really keen sense of what is needed for amputees now and what is possible to put into a system that won't have an inordinate amount of complexity," says Dr. Englehart, who works at the University of New Brunswick's Institute of Biomedical Engineering in Fredericton.
His work on the fully bionic arm was funded by an agency that does research for the U.S. military. As an intermediate step, the international team came up with Proto 2, an artificial arm that can move in 27 different ways. Amputees use nerves that have been transplanted or "redirected" by surgeons to help control it.
His current project, the new artificial hand, builds on that research, and aims to dramatically improve the quality of life of many amputees. Most amputees - about 60 per cent - lose a portion of their limb below the elbow.
This new hand will be far more dexterous than any currently available, he says. It will have realistic dynamic movement, so that it will look more like a real hand when it is moving.
The researchers at UNB work closely with patients at a clinic, says Dr. Englehart, which helps them understand what's needed. As with home electronics, consumers don't want a system that is difficult to use, he says.
The heart of the project is a human-machine interface that will decode the signals from the muscles of the lower arm and tell the prosthetic hand exactly what to do. "They will simply contract their muscle in the same manner as they would move an intact limb, he says.
He says it is important that it be affordable. The hand will likely cost about $10,000, less than a tenth of Proto 2, which is still an experimental device and not manufactured commercially.
"It will contain all the essential components of the more ambitious projects but at a price that insurance companies can get behind," he says.
He and his colleagues are 3 1/2 years into the five-year artificial hand project, which is funded by a $2.9-million grant from the Atlantic Innovation Fund.
They hope to start the first trials by the end of the year and have the system ready to go in 2012.
Dr. Englehart collaborates closely with Todd Kuiken, director of amputee services at the Rehabilitation Institute of Chicago.
They worked together to help Jesse Sullivan, who in 2001 lost both of his arms at the shoulder when he touched a high tension wire while working as a lineman for a Tennessee power company. It carried 7,400 volts of electricity; both Mr. Sullivan's arms suffered severe electrical burns and had to be amputated.
Dr. Kuiken took the main nerves that had connected to Mr. Sullivan's left arm and fastened them to a muscle on his chest. After about six months, the nerves grew into the muscle. A computer in his artificial limb picked up signals from those nerves to his absent arm and hand.
Dr. Englehart helped come up with the algorithms that would translate those signals into commands for the artificial arm.
It was technically challenging. In an intact limb, the nerves to the arm are bundled together at the shoulder, then branch out lower down and connect with specific muscles.
The transplanted nerves had been cut off near the shoulder and were sending out many different signals at once. Dr. Englehart and Dr. Kuiken did experiments with Mr. Sullivan to tease out what those signals meant.
The next step was to build software for the artificial limb that would accurately interpret whether the patient wanted to open or close a hand, give a thumbs up or a thumbs down.
Other researchers worked on sending signals from the hand, through the transplanted nerves, to Mr. Sullivan's brain so that he could sense hot and cold and feel enough sensation from his prosthetic hand to pick up eggs without breaking them.
It was the first time that a nerve-muscle graft was used to control an artificial limb.
The ground-breaking work caught the attention of DARPA, the Defense Advanced Research Projects Agency. It does research for the U.S. military and wanted to jump-start the production of neurally-controlled artificial limbs.
Dr. Englehart and his colleagues at UNB had a substantial role in developing an embedded computer that could interpret muscle activity and relay information about how to move the prosthesis.
Although the bionic arm project officially ended in 2009, a number of teams are continuing to work toward creating one. A major obstacle is developing a neural interface - a way to intercept signals from the brain that would control an arm or leg - that the body won't reject, says Dr. Englehart.
About 40 people have had the same kind of surgery as Mr. Sullivan, which, when coupled with the algorithms developed at UNB, allows them to use existing prostheses more naturally.
But there remain technical issues. Electrodes on the skin are needed to send and receive messages from the transplanted nerves. If a patient sweats too much, those electrodes can stop working. Transplantable electrodes, which would solve the problem, are on the horizon.
This work is also being applied to other limbs. One of Dr. Englehart's former graduate students, Levi Hargrove, is now working with Dr. Kuiken on applying their findings to artificial legs, and Dr. Englehart is a collaborator on the project.
The project also involves decoding muscle signals, he says. But the prostheses must be more powerful, because they have to bear the weight of the user. Mistakes in decoding the neural information are more costly, he says.
"You fall down."