Trends in powered upper limb prostheses
I. Kato *
Introduction
As Prof. Tomovic pointed out at the 3rd ETAN Symposium in 1969, realization of multi-functional mechanisms and development of multilevel control methods were the main subjects areas of interest in the field of powered upper limb prostheses in the latter half of the 1960's. In addition, the use of myoelectric potential as the source of control signals and utilization of electric stimulation for sensory feedback became a common method in the 5th ETAN Symposium of 1975.
The utilization of myoelectric potential as the source of control signals was reported for the first time by Battye et al. in 1955 and the use of pulse electric stimulation as the feed-back signal by Kato et al. in 19701. While thees methods utilize a surface electrode, new attempts such as the direct insertion of an electrode into a nerve have been started. These will be described later.
Recently, a manipulator for the disabled was developed with multiple degrees of freedom. This can be placed on the working desk of an amputee requiring a prosthesis or placed on the table of a motorized wheelchair for patients with spinal cord injury requiring an orthosis. The patients can use the manipulator, which is driven by some signal, to accomplish their work.
Forearm prosthesis with one degree of freedom
The electric forearm prosthesis has already been introduced into practical use in many countries. The apparatus developed jointly by the Veterans Administration Prosthetic Centre (VAPC), and Northwestern University (NU) became commercially available as the Fidelity Hand (VA-NU) in 1973. The mechanical characteristics of this prosthesis are the variable grasping force of the finger up to 5-5 kg, the spontaneous release of the grasp through dorsiflexion of the middle phalanx of the index and middle fingers in response to an overload above the pre-set value, and the packing of the battery within the wrist.
These mechanisms were recently employed in MYOMOT of Viennatone (Type MM4S). In MYOMOT, the conventional wrist band type battery was used in addition to the built-in type battery described above.
In Japan, the Waseda-Imasen-Myoelectric (WIME) hand-4P (Fig. 1 ) (Kato et al. 1970) revised by Imasen Electric Industry Company Limited is undergoing field tests on about 25 amputees at the National Rehabilitation Centre for Handicapped, the Labour Accident Pros-thetics and Orthotics Centre, the Tokyo Metropolitan Prosthetic and Orthotic Research Institute, and the Hyogo Rehabilitation Centre (Kato et al.,1976,2 1978), (Fig. 2 ).
After more than two years of clinical testing sponsored by the Research Development Corporation of Japan, it is anticipated that it will be introduced into practical use within one year. The mechanism and control system of this apparatus are quite different from those of the forearm prostheses already commercially available from Europe and America. The mechanical part is provided with five fingers of adaptive type, each with basal and middle phalanges providing a grasp with 5 fingers and pinch with 3 fingers made possible by mechanical flip-flop. An ordinary electric forearm prosthesis measures the difference in myoelectric potentials of 2 channels and utilizes its polarity as the signal for opening and closing of the fingers. The WIME hand system utilizes the preferred method of generation of 2 channel signals, or the signal reaching the threshold determines the pattern. The Waseda University, the Mechanical Engineering Laboratory, and the Japan Bicycle Technical Centre are also co-operating in this project.
Forearm prosthesis with 3 degrees of freedom
Regarding the forearm prosthesis with 3 degrees of freedom, the final report of SVEN project-1 forearm prosthesis started in 1965 was submitted in 1973 (Hägg et al. 1973, Herberts et al. 1973, Almström 1977). Adaptive grasping, pronation and supination of the wrist, dorsiflexion of the hand, and lateral flexion of the wrist (manual movement) represented the available free movement. A similar mechanism was tentatively produced by Karas (1975) in Austria.
The National Defence Research Institute (FOA) took part in the production of the mechanical portion of the SVEN Hand. Its control system was developed in the Chalmers University of Technology. It is my understanding that a test of its application was conducted on several cases.
In Japan, a hydraulic type hand prosthesis utilizing myoelectric potential (Fig. 3 ) is being developed by the WASEDA University, the Tokyo Metropolitan Prosthetic and Orthotic Research Institute, the Tokyo Metropolitan Institute of Gerontology, Mitsubishi Metal Corporation and Kayaba Industry Company Ltd. This model use a new type hydraulic actuator, the Rotary Servo Actuator (RSA) ([link4} ), and has 3 degrees of freedom; opening-closing of fingers, dorsiflexion of the hand, and pronation-supination of the wrist. RSA is a torque amplifier with an internal mechanical feedback pathway, incorporating a rotary spool valve of a high degree of precision within the rotary actuator. The ultra-small hydraulic source (Fig. 5 ) for driving the device consists of a micropump, an electric motor, and an oil reservoir incorporated into one unit.
A system is introduced in which the space distribution information of the multichannel EMG is utilized for pattern recognition for the co-ordinated control of pronation and supination, dorsiflexion of the hand and individual control of opening and closing of fingers. This control system is similar to the Chalmers system (Ichikawa et al., 1975). Clinical testing of this system was recently started (Kato et al., 1976").
Developments
The primary difficulty in the control of a powered hand prosthesis is how to obtain the control signal. This problem concerns the portion of peripheral nerve and the pulse transformation methods of EMG and its interpretation.
For this reason, attempts were made to obtain the control signal by the use of an implanted electrode. A new electrode was constructed and tested on laboratory rabbits for 244 days (Luka, 1975). The new electrode was sutured in association with the longitudinal axis of the neural sheath. This was connected with the extracorporeal amplifier via a carbon glass wire. For this new electrode, a high quality carbon (bio-carbon) developed by NASA was used and connected with the external wire via a SmCo magnet (REC, 1973). This is called Biosnap. Within the next three years, it is planned that control signals will be picked up from the musculocutaneous, radial, median and ulnar nerves by this neural electrode method.
The transmission of the state of control of the prosthesis to the user in a feedback mechanism remains another problem; that is, how to construct the portion equivalent to peripheral nerve. In the powered hand prosthesis in current use, the visual pathway is mainly utilized for such a feedback mechanism, as well as the change of sound in the driving unit and detection of the change of compression at the point of contact with the socket. The use of visual sensation involves the action of the central nervous system during the manipulation of the prosthesis, leading to enhanced work load on the side of the user. With the use of prostheses utilizing the internal power source, the state of loading is directly transmitted to the contralateral shoulder or the muscle subjected to cine-plasty representing the power source. Thus, the magnitude of loading may be sensed unconsciously via the proprioceptors at these sites. For this reason, the amputee frequently prefers the system with internal power source to the system driven by a motor. For the propagation of powered prostheses, provision of the "peripheral neural pathway" is considered to be one of the requirements. Transmission of the feedback signals as pressure by a portion of the prosthesis to the skin of the user has been attempted by electric or mechanical stimulation (Kato et al. 1970, Mann 1970).
In the new attempts by Reswick et al. (1975), the idea is further advanced by directly stimulating the remaining nerve in a neuro-electrode system. The information on the pressure and the position of the hand taken up from each sensor is transmitted percutaneously into the body as a pulsed electric stimulation via a Biosnap. A coil electrode was inserted about 10 mm into the remaining sensory nerve to transmit the informa-tion to the centre.
The insertion of this electrode was already attempted in amputees by attaching the sensor to a hook of the APRL type. Similar experiments have already been started using VA-NU myoelectric hand prosthesis.
The author was invited to the 5th INTERBOR meeting in Paris in the spring of 1972 and had an opportunity of observing the exhibit of a hand prosthesis equipped with one mini-motor to each of the basal phalanges of 5 fingers. The prosthesis was designed by Barrachina et al. of the Rehabilitation Centre, French Veterans Administration. This hand prosthesis has been repeatedly revised by the group headed by Rabischong, and is known as a téleméchanique hand. Hill et al. (1975) suggested a system of incorporating various kinds of action control circuit as modular cards, in order to select the desired action pattern by inserting the proper card.
Heer (1975) started a project of applying tele-operator techniques accumulated over many years at NASA to an instrument for the disabled in collaboration with VAPC. As the first step, a manipulator (Fig. 7 ) with 6 degrees of freedom was placed on the table of an electrically driven chair and attempts were made to control this with a jaw-controlled bar and by voice through a minicomputer equipped to discriminate 35 words such as "up", "down", "right" and "left". This reached almost a practical stage. Though no especially new technique is used, this idea suggests a future direction in developing a prosthesis for the disabled.
Method of evaluation
In the case of an ordinary industrial instrument, developmental research is conducted via basic research and exploratory research. Whenever it is necessary, instrument tests are conducted for practical use. In the rehabilitation instrument for the disabled, however, these processes are insufficient and a field test for practical use is indispensable after the developmental research.
Field testing is performed in 3 stages. Mechanical tests are not so different from those for ordinary instruments. The focus of field testing rests in human-machine tests and psycho-sociological tests. The objective evaluation of this system on a human being is very difficult since it involves an evaluation of many facets of an individual.
As a method of human-machine testing, Activity in Daily Life (ADL) has already been adopted in the medical field. Since ADL indicates a series of repeated physical activities fundamental for independent living of an individual, this test alone gives insufficient data for the evaluation of adaptability to occupational life. In ADL testing, moreover, the subjective view of the evaluator considerably influences the results.
As an objective evaluation of human-machine testing, functional testing using M-series signals or abstract action testing has been suggested (Kato et al., 1976). The former is the method of evaluation of human-machine systems using binary random signals in order to avoid ambiguous evaluation due to non-linearity and learning ability of human beings. The latter is the method of objective evaluation of function after abstracting and decomposing general action including occupational action into fundamental components.
In practice, these methods of evaluation are used in combination to carefully determine the suitability and occupational fitness of instruments.
Psycho-sociological testing evaluates the human-machine system described above with reference to living and occupational environment. Psychological aspects of the user are emphasized and items such as adaptation to society and changes of consciousness and behaviour are selected for survey.
As to the overall evaluation of these tests, the degree of mechanical restoration of the lost function becomes the starting point for the guidance, training, and selected introduction in occupation rehabilitation.
In Japan, this method of evaluation was applied to the field test for the WIME Hand and a summary of the results is in press (Kato et al., 1978).
References:
Almström, C. (1977). Myoelectric control of multi-functional hand prosthesis-contributions to the pattern recognition approach, to signal acquisition and to clinical evaluation. Technical report No. 79, Chalmers University of Technology.
Battye, C. K., nightingale, A. and whillis, J. (1955). The use of myoelectric currents in the operation of prostheses. J. Bone and Jt. Surg., 37B, 506-510.
Hägg, G. et al. (1973). SVENProject 1-Electrically controlled hand prosthesis. Final report FOA 2.
Heer, H. (1975). Voice controlled adaptive manipulator and mobility systems for the severely handicapped. Preprint of 2nd Conf. on remotely manned systems.
Herberts, P. et al. (1973). Hand prosthesis control via myoelectric patterns, Acta Ortho. Scand., 44 (4-5), 389-109.
Hill, J. W. et al. (1975). Control system for a multifunction prosthetic hand. Proc. of 5th Int. Symp. on External Control of Human Extremities ETAN, 469-478. Dubrovnik.
Ichikawa, K. et al. (1975). Analysis of myoelectric potential as control signal of hand prosthesis. Biomechanisms 3 82-90 (in Japanese). karas, von W. (1975). Die Weiterentwicklung der Wiener Adaptivhand, Orthopädie-Tecknik, 26:4, 66-68.
Kato, I. et al. (1970).1 Human cognitional ability for electric stimulation signals. Proc. of 3rd Int. Symp. on External Control of Human Extremities ETAN, 69-84. Dubrovnik.
Kato,I.et al. (1970).2 Multi-functional myoelectrical hand prosthesis with pressure sensory feedback system-WASEDA Hand-4P Proc. of 3rd Int. Symp. on External Control of Human Extremities ETAN, 155-170. Dubrovnik.
Kato,I..et al. (1976).1 Electro-hydraulically operated multifunctional fore-arm prosthesis-Waseda hand 9H3-Foundation for Research on Medical and Biological Engineering (in Japanese).
Kato, I. et al. (1976).2 Evaluation of rehabilitation devices. System Technology Report 50-12-2 Kikaishinko-Kyokai (in Japanese).
Kato, I. et al. (1978) . The evaluation method of rehabilitation devices. Field testing of powered fore-arm prosthesis, WIME Hand. Proc. of 6th Int. Symp. on External Control of Human Extremities ETAN. Dubrovnik. (In press).
Luca, C. J. (1975). Consideration for using the nerve signal as a control source for above-elbow prosthesis. Proc. of 5th Int. Symp. on External Control of Human Extremities ETAN, 101-112. Dubrovnik.
Mann, R. W. et al. (1970). Kinesthetic sensing for the EMG controlled "Boston Arm". Proc. of 3rd Int. Symp. on External Control of Human Extremities ETAN, 231-243. Dubrovnik.
REC at Rancho Los Amigos Hospital. (1973). Annual Reports of Progress 16 and 58-61.
Reswick, J. et al. (1975). Sensory feedback prosthesis using intraneural electrodes. Proc. of 5th Int. Symp. on External Control of Human Extremities ETAN. Dubrovnik.
|
ISPO Home