The term “bionic” once had a purely scientific meaning. Now, thanks to pop culture like The Six-Million Dollar Man, what immediately comes to mind are images of someone leaping over buildings, lifting cars, and performing other feats of superhuman strength, associated with the words: “We can rebuild him; we have the technology. … Better than he was before. Better. Stronger. Faster.” It’s like trying to listen to Rossini’s William Tell Overture without thinking of the Lone Ranger.
While it is highly unlikely that bionics will ever give us superhuman powers, it’s true that there have been dramatic improvements in prosthetic technology over the past 30 years. What once seemed like science fiction is rapidly becoming reality. Although even the highest state-of-theart prostheses only approach the replication of actual limbs, the exciting union of bionics and prosthetics offers the promise of technologies to come that will greatly improve mobility, energy efficiency and gait patterns.
Sensors and Sensibilities: Better Prostheses Through Enhanced AI
The key to future prosthetics, experts say, lies in enhancing artificial intelligence, or AI. The concept of integrating computer technology into prosthetic designs isn’t new. AI is already used to monitor self-adjusting knees; Otto Bock, Seattle Systems and Endolite have all made progress in this area. Artificial hands are now being manufactured with sensors in the fingertips that can monitor the force of the grip.
There are many other possible applications. For example, research is being conducted on sensors embedded in socket liners, which hold the prosthesis to the residual limb, that will automatically adapt to fluctuation in body volume. Self adjusting sockets could make the device more comfortable and prevent sores, bruises and other complications. Similar sensor technology could also enable artificial feet to stiffen or relax to adapt to changing resistance and speed, making it possible to use the same device for a variety of activities.
The biggest obstacle yet to be overcome is replicating the complexity of the entire range of limb movement with existing technology. While great strides have been made, there are a number of interrelated factors, such as power, response speed, range of motion, and weight and durability, which still must be resolved. The ultimate goal in solving these problems is the complete integration of mind, body and machine.
The Ghost in the Machine
In 2000, Dr. Miguel Nicolelis, a neurobiologist at Duke University Medical Center, trained a monkey to move a robotic arm using thoughts transmitted through electrodes implanted in its brain. Recent experiments show the technique may work in humans as well.
The key is to identify the neurons that are activated when someone consciously thinks about a movement and then performs it. Studies show that these brain cells remain active even in amputees. Nicolelis took advantage of a series of brain surgeries performed on 11 patients with Parkinson’s disease, an incurable disease marked by the destruction of certain brain cells. These operations are routinely performed with the use of deep brain stimulators to help counteract the tremors associated with Parkinson’s disease. Nicolelis made the most of these opportunities by adding a simple manual task: The patients were asked to play a video game during the operation.
To find the best place to put the stimulators, surgeons implanted temporary arrays of microelectrodes. The patients were awake during surgery so they could guide the surgeon. As the patients played the game, the electrodes sent their brain signals to a computer, which analyzed the signals and matched them with the hand movements used during the video game. The signals were then compiled into a neuronal model that could predict the signals needed to perform the actions required by the game.
While the monkeys had wires implanted in their skulls that were connected to an external device, Nicolelis’ team has designed a wireless electrode array that could be implanted permanently. Some obvious applications of such technology would be a thought-controlled robotic arm, electric wheelchair, TV, computer or prosthesis. Dr. Josý Millýn at the Dalle Molle Institute in Switzerland has come up with a similar system. Instead of surgical implants, however, a cap studded with electrodes monitors brain activity through the scalp. Early trials using a steerable robot indicate that it is as easy to control the robot by thought as it is manually.
A web of wires sends the data to a computer that analyzes the brain’s activity and, using a wireless link, passes on any commands it spots to the robot. The computer program uses the fact that the desire to move in a particular direction generates a unique pattern of brain activity. It can tell which command the user is thinking of by identifying the neural signal associated with that command.
Traditional prosthetic arms are bodypowered, relying upon body movements to manipulate cables that control the prosthesis. Myoelectric prostheses are controlled by electrical signals transmitted from underlying muscles to the surface of the skin, which are then amplified and sent to microprocessors that operate motors in the joints and hands. Recent advances in the technology are featured in the new SensorHand® Speed by Otto Bock and the Utah Arm 3, soon to be released by Motion Control.
The SensorHand Speed is quieter than the original SensorHand, more than twice as fast, and more responsive due to enhanced signal processing and software. The “Autograsp” feature senses when an object held in the hand requires more grip force and automatically adjusts tension, such as when the hand is holding a glass that is being filled with water.
The Utah Arm 3 is in the final stages of testing and is expected to be released to the public market by 2005. Designed for above-elbow amputees, it contains two microprocessors that enable simultaneous elbow and wrist or hand movements. Another upgrade is the ability to alter the power settings; for example, the user can amplify the signal pickup to compensate for lower neural output when the muscles are fatigued.
Sometimes the injuries that cause an amputation also damage the muscles that would typically be used to control an electronic prosthesis. The Rehabilitation Institute of Chicago (RIC), a leading physical rehabilitation hospital in the U.S., recently completed the successful rehabilitation of the world’s first recipient of a new nerve/muscle graft procedure to control a myoelectric prosthesis. The technique, researched and developed by Dr. Todd Kuiken, MD, PhD, director of amputee services, allows an arm prosthesis to be operated by different portions of the user’s remnant muscles.
On May 21, 2001, Jesse Sullivan of Dayton, Tennessee, touched a live wire on the job as a power lineman. His electrical burns were so severe that they required both of his arms to be amputated at the shoulder. Sullivan had limited myoelectric control sites and was fitted with a body-powered prosthesis for the right arm.
Dr. Kuiken proposed transplanting living nerves from Sullivan’s left shoulder into his left pectoral (chest) muscle. Six months later, after the nerves had grown and spread into the muscle, Dr. Kuiken was able to detect myoelectric signals from them. RIC’s prosthetics department then designed and built a “cuff” with electrodes that pick up and translate the signals to control the prosthesis. When Sullivan thinks about closing his hand, the nerve that once made the hand close now causes part of his chest muscle to contract. Sensors over that muscle then tell the hand to close via tiny connecting wires.
“We developed and tested this procedure for years; this is 1920s surgery for a 21st century application.,” says Dr. Kuiken. “What’s really novel about this is not so much the surgical technique, but the reason for doing the surgery and using it to help control artificial limbs and make them work better.” The technique is limited to amputated arms for now, although it is hoped that it can be applied to legs as well, eventually.
The Dextra hand, invented by Dr. William Craelius of Rutgers University (November/December 2002 inMotion), takes the concept one step further. It is the first device of its kind to enable a person to use existing nerve pathways to control individual, computer-driven mechanical fingers. Dextra has been demonstrated in such complex activities as typing and piano playing. Although the level of dexterity is below that which would be required to play Flight of the Bumblebee, Craelius believes “bionic technologies can be adapted for restoring some degree of almost any lost function.”
For example, he cites a tiny, wireless implant developed at UCLA by a team led by Dr. Gerald Loeb. This can be injected under the skin to transmit neural signals to bionic devices. According to Craelius, while it may require more than 1,000 connections between the brain and bionic devices to communicate the data, it is probably achievable using existing technology, even if most of the necessary computer processing would have to be done outside the body. At the current rate of increasing computing power, speed and miniaturization, however, Craelius says that the entire processing system may be able to be contained internally, in the brain or elsewhere in the body, in the next 10 years.
The Dextra technology is now being used in a collaborative study with Dr. Grigore Burdea of Rutgers to evaluate how it may benefit amputees through the use of virtual reality. Wearing a cuff that picks up signals from the user’s arm, the user controls a virtual onscreen hand to perform dexterity exercises on a computer. Preliminary tests indicate several possible uses for and benefits of this technology. It might be useful as a tool for assessing amputees and training them to use a mechanical hand; it might be integrated into future designs to improve the function of prosthetic arms and hands; and, in the case of lower-limb amputees, it might be used to help to improve their gait. Perhaps most interesting of all, virtual reality may help amputees cope with phantom pain by enabling them to visualize and move a virtual lost limb into a less painful position.
The problem with traditional prostheses has always been that the wearer does most of the work, particularly when walking uphill or over uneven terrain. That may soon become a thing of the past.
Iceland’s Össur Total Prosthetic Solutions has signed a deal with Victhom Human Bionics Inc., Canada, to manufacture and distribute what has been called the world’s first bionic leg, a motorized, prosthetic system for transfemoral amputees. Victhom’s leg technology is different because it works with, not against, the wearer. Unlike passive lower limbs, Victhom’s knee is capable of both active flexion and extension, which makes it more energy-efficient and enables a more natural gait.
Sensors in the amputee’s shoes send signals to the knee’s computer, which reproduces the appropriate walking pattern. Whether you want to climb stairs, walk quickly or slowly, sit or stand, it recognizes it in real time.
The new bionic limb is still undergoing testing and may be on the market by early 2005. Victhom is now working on a motorized ankle and artificial muscles, as well as a neural implant to replace the bionic leg’s external sensors.
Seattle Systems, in conjunction with Sandia National Laboratories, is still working on a Smart Integrated Lower Limb (SILL). The final version will incorporate multiple sensors that feed data on pressure, position, and speed to a central processor that will control the knee as well as the ankle, foot and socket. The leg socket will adjust to the changing diameter of the wearer’s residual limb over the course of a day. Pressure sensors in the foot will deliver a mild buzz to electrodes attached to the residual limb. Using these cues, amputees will be able to train their limb to “feel” their prosthetic foot as it hits the ground.
The basic concept of the SILL is common to the schools of thought behind all bionics: An artificial limb, like a healthy one, should function as a unit, rather than a group of parts. This is the definition of bionic technology, in the truest, simplest sense. Total integration – a direct link between human and machine.