Clinical Experience With A Myoelectric Prosthesis
Andrew L. Lippay, B.Eng. E. David Sherman, M.D., F.A.C.P., Director of Research; G. Gingras, M.D., F.R.C.P.(C), Executive Director.
When physical disability deprives the human body of essential functions, orthotists and prosthetists apply their art to restore missing limbs, to strengthen and reactivate flail extremities, and to re-establish cosmetic completeness. However, to be functionally useful the artificial limb or orthotic appliance must be activated. As the extent of disability increases, so does the need for more and more artificial functions, while the remaining body segments capable of providing motive power progressively decrease.
To meet this challenge, artificial effectors or actuators, with portable sources of energy external to the body, are needed. Suitable structures and terminal devices must then be developed to produce useful movement and functions. Most important of all, control systems must be devised to connect and match the human system to the mechanism, so that the latter will obey the will and commands of its operator. Ideally, this man-machine communication should be bidirectional: the control system should provide information feedback to the operator, as the human nervous system advises the brain of conditions existing in various parts of the body.
Electricity and compressed gas are two convenient forms of portable energy. Each has its advantages and disadvantages, but impending technological developments should eliminate most of the drawbacks to both systems in the near future. True engineering efficiency may often have to be sacrificed for the sake of reliability and patient convenience. Individual applications require great flexibility and adaptability of components, pointing toward block-by-block or modular construction systems. Compared to conventional types, the cost of externally powered prostheses is high at present, but in the light of their wide perspective and rehabilitative potentials, cost considerations should never be a limiting factor.
Control of External Power
In the conventional orthotic or prosthetic appliance, control is effected by readjustment of the neuromuscular channel associated with the body segment that activates the device. For example, humeral flexion may be used to provide terminal-device operation in the above-elbow-prosthesis. In externally powered systems, the operator is required to perform a control function only; little or no physical exertion is involved. As an illustration, a force exerted on the accelerator pedal controls the speed of the automobile, but the driver's effort is infinitesimal compared to the mechanical power developed by the engine in driving the car forward, using the energy stored in the fuel.
Machine control may be achieved by means of switches, valves, pedals, and a great variety of automatic devices accepting command signals in one form or another. In rehabilitative man-machine combinations, any function of the body may be used for control, provided that such activity is under voluntary control, occurs with reasonable speed, is independent of other essential functions, and involves a body site where a detecting device (transducer or sensor) may be placed and maintained indefinitely.
While a human operator can adapt to almost any type of machine control, there is a limit to the number of discrete activities he can monitor simultaneously. This limit may be extended by training, but excessively complex functions may remain beyond the capacity of some patients. On the other hand, functions may be preprogrammed to be completely automatic, with the operator being required only to initiate and to stop the program. Too much automation, however, tends to robotize the patient, and it is doubtful whether such systems would be accepted as part of the body image. The existence of mechanical limitations and slow response in a device will produce severe patient frustration, discouraging its use. If a control system is to establish true man-machine communications, its design must be compatible with both.
Myoelectric Potentials as Control Signals
The existence and the nature of muscle action potentials have been known to biological and clinical investigators for more than a century. The transistor and modern electronics have permitted the development of small, portable myoelectric amplifiers and control circuitry capable of controlling electric actuators and devices according to the intensity of the electrical byproduct of muscular activity. While the energy level of the action potentials is much too low to power the devices, a satisfactory control signal may be derived from them and with adequate amplification can be used to regulate the flow of external energy.
Although many myoelectric control problems remain unsolved, the practicability of such systems has been proved both in the laboratory and in the clinic. Thus a completely new field may be opening up as the possibility of using single motor unit pulses for control attracts the attention of specialists in the field. With the lack of control sites available for extensive prosthetic applications in severely handicapped individuals, myoelectrics may well prove to be the sine qua non of such systems, providing an intimate contact between man and machine. In the more distant future an even closer connection may be established between the central nervous system and artificial devices. The electrical nature of nervous signals suggests that combining human and electronic systems will yield interesting and useful results.
The Soviet Myoelectric Prosthesis
The active elements in the Soviet myoelectric prosthesis (Fig. 1 ) are a high-speed electric motor and a gear drive located in the metacarpal area of a hollow plastic hand. The inter-phalangeal joints of the phalanges are rigid but slightly flexed. They are hinged at the metacarpophalangeal joint. The thumb is straight and hinged to the handpiece in an opponent position. The rotation of the motor in one direction or the other is converted through a gear reducer and lead-screw combination to produce linear motion in two directions. This movement is linked to the index and long fingers and the thumb, causing them to open or close in a three-point pinch.
The hand opens or closes in 0.8 seconds, slower at ambient temperatures below freezing. The final pinch force is approximately 1.5 kg, with a maximum of 2 kg, or roughly 4 pounds.
The motor is energized from a 13.75-volt battery, which also powers the control amplifier. Current may be switched on through two small relays, producing rotation in either direction. Each of the relays is activated by a control channel consisting of a muscle under voluntary command, a set of surface contact electrodes, a sensitive myoelectric amplifier, an integrating block, and a power amplifier. The latter energizes the relay directly, while blocking the other channel to prevent simultaneous operation.
Whenever the control (muscle) activity is sufficiently high, the corresponding relay will switch on and the motor will run in the appropriate direction at a constant speed until an object is encountered, loading the hand, or until the mechanical stop prevents further movement at the end of the cycle. To some degree the force of the grip may be graded by closing the hand in small increments and observing the deformation of the object held, or by estimating the force from the sound of the motor. This gradation, however, is difficult with hard or brittle objects.
The remainder of the prosthesis consists of a passive wrist unit which permits pronation-supination only, and a standard forearm socket with optional leather harnessing. The hand is covered with a cosmetic glove, the surface friction of which increases the effectiveness of the grip.
In the original design the battery and control amplifier were carried in two small pockets on a belt. The electrodes were spring-loaded and contained an abrasive paste in a small cup. A separate neutral electrode was applied to the skin of the forearm.
In Montreal the original Soviet equipment has been modified in several respects (Fig. 2 ). On the basis of clinical experience, all wiring has now been incorporated in the structure of the forearm socket; Canadian and American hardware has been substituted to improve the reliability and convenience of the device and to increase the patient's comfort. A versatile new battery package utilizing a flexible plastic enclosure was developed. A Hosmer wrist unit (FM-100) was adapted to the hand to provide a positive lock and fine rotational adjustment. A set of three solid stainless-steel electrodes has eliminated the need for a separate neutral point and for an electrode paste.
Microminiature electronics are presently employed to produce a control package containing the electrodes, amplifiers, and relays, which can be fitted into the prosthetic socket without unsightly protrusions. A universal wrist unit is being adapted to provide passive adjustment of wrist flexion and extension, as well as pronation-supination.
Training Methods and Instrumentation
In developing the two control sites required by the Soviet system it is first necessary to locate two convenient muscles with good strength and control. The forearm is surveyed with a small indicator instrument to determine the optimum sites for the location of the electrodes. Wherever possible, a finger extensor muscle is chosen to open the hand and a flexor muscle to close it. At present we are also investigating trunk muscles as prospective control sites for amputations above the elbow and higher.
The patient is shown the activity of his muscles on multichannel indicators (Fig. 3 ), which seem to be more meaningful to him than the oscillo-scopic trace. In some cases, such as that of a blind patient, an audible display has been used as a guide to develop discrete operation of the control muscles (Fig. 4 ).
The patient's interest and motivation improve considerably when, for the first time, he is allowed to operate a demonstrator hand, even though the latter is only a model on the table in front of him. In general, a patient with the intelligence to adapt to unusual situations, and with sufficient self-discipline to follow instructions, will learn to control the basic movement of the hand in minutes rather than hours. Three or four one-hour sessions spaced not more than two days apart are usually sufficient to establish the location of the electrodes. The patient can then be referred for prosthetic fitting and occupational therapy training.
The amputee's mental picture of his missing hand (phantom sensation) is used to good advantage in the preliminary training. Should construction of the prosthesis be delayed, practice sessions should be continued to maintain the control pattern. Physical therapy may be used to improve the strength of control signals through exercise.
The fitting and training of patients with bilateral amputations present no greater difficulty than is the case with unilateral ones. In a few hours of practice they learn to operate the two hands simultaneously and independently. Operation of the prosthesis soon becomes automatic, almost subconscious, and is usually associated with the activity of the phantom image.
The Prosthesis in Everyday Use
Fourteen adult below-elbow amputees, including one female, have been successfully equipped with a myoelectric prosthesis. An additional case had to be abandoned due to lack of interest and motivation. The ages of the patients ranged from 17 to 66 years, the amputation history from a few months to 28 years. Three are bilateral cases. One unilateral case is also totally blind, and one bilateral case has only one eye.
All of the above installations have been successful technically, although in one case the functional improvement over other prostheses and the degree of satisfaction achieved are rather doubtful.
Some of our units have been in operation on patients for 18 months or more. After a brief initial period, maintenance problems have been few and far between and have consisted of loose connections, broken wires, etc. In most cases the defect was corrected in a matter of hours, the main inconvenience being the delay due to mailing and shipping, clearing customs, etc. Operating costs have been negligible .
The mechanical drive has presented no difficulties.
Two of the bilateral cases suffered major damage to their hands, such as breaking their thumbs, etc., in accidents, which indicates vigorous use of the prosthesis.
The service life of the glove varies from one to six months, depending upon the amputee. Some gloves of American origin have been tried, but they did not fit satisfactorily. The wearing qualities of the United States item, however, are superior.
The new battery package sustains operations for up to 15 hours between charges, with no malfunctions reported so far. A charge indicator has been requested by several of the users, but the available commercial devices are not suitable for home use. However, the present batteries are not excessively sensitive to overcharging and similar mild abuse.
Great emphasis is placed on simple and independent application of the prosthesis. Where a Muenster fitting is not practicable, a triceps cuff or shoulder harness is employed for unilateral cases. Bilateral equipment is connected with a light shoulder harness, which also supports the batteries on a belt. A conductive paste, Aquasonic 100, is applied indiscriminately over the skin in the electrode area to establish contact quickly. However, once conditions have been stabilized in the arm socket, the electrodes require no further attention. In cases of heavy perspiration, electrode contact has been successfully established through a single layer of stump stockinette.
Several of the amputees have reported involuntary operation of their hands on board airliners and in certain areas of the city. The source of the trouble is probably a concentration of high-frequency electromagnetic disturbance, such as is created by aircraft radio equipment and microwave communications. We are currently attempting to establish the conditions causing the disturbance through discussions with an airline.
The first nine patients fitted were sent a three-part questionnaire approximately six months after their discharge. The three sections covered the use and acceptance of the prosthesis, comfort and convenience of the fit, and mechanical and electrical performance. All patients completed and returned the questionnaires.
The answers dealing with the actual use of the arm varied fairly widely, ranging from use on Sundays and special occasions only to wear during the entire waking day. With one exception, all the patients had previously used a conventional type of prosthesis, and many still used the other prosthesis, particularly for manual or heavy work. However, some used their hands exclusively and had become dependent on them. They reported that members of their families and their friends thought well of the arm and commented favorably on its cosmetic appearance and its functional ability. With few exceptions, the answers indicated that the arm was reasonably comfortable and that no pressure sores had developed. However, the myoelectric arm was also reported as heavy and awkward, presumably in comparison with the conventional type formerly worn. All respondents mentioned the difficulty they had experienced in pushing the arm through a sleeve, and some were not satisfied with the color and appearance of the glove.
On the electrical side, few amputees had experienced stoppages of any consequence, they had little difficulty in establishing contact, weather or body conditions had little effect on the control of the arm, and charging the battery had caused them no inconvenience. All reported, however, that the arm was subject to external influences.
Andrew L. Lippay, B.Eng. is the Engineering Consultant Department of Research Rehabilitation Institute of Montreal Montreal, Quebec