Appendix B: The M.I.T. Experience
Robert W. Mann
Work supported by Social and Rehabilitation Service, HEW; and Liberty Mutual Insurance Company.
The research and development effort associated with the "Boston Arm" recognized from the outset that the substitution of electromyographic control for the traditional cable would deprive the amputee of kinesthetic force and position information on elbow flexion. This feedback is derived from "sense-of-effort" in the body motions used to actuate the cable. To replicate force feedback in the EMG elbow, strain-gaged elements were introduced into the prosthesis. Their output was subtracted from the processed electromyographic signal from the dysfunctional but relevant biceps-triceps muscle pair. The difference signal drove the proportional servo-control of the elbow (Fig. 1 ). Thus increased terminal device load or acceleration force induced a physiological increase in muscle contraction; intact muscle and tendon sense organs in the biceps and triceps "informed" the amputee of "sense-of-effort"1.
Sensory feedback of elbow angle was demonstrated and evaluated using an electromechanical presentation of the cutaneous "phantom" sensation via two vibrators on the upper arm socket stroking the amputee's stump2,3. An objective, computer-controlled, lengthy experimental sequence compared the EMG-controlled prosthesis, with and without the cutaneous position feedback, with and without audition, versus the amputee's control of his cable-operated prosthesis and his contralateral normal arm, all without vision (Fig. 2 ). The results demonstrated that this cutaneous feedback at the least replicated, and in some cases achieved superior performance to, the position kinesthesia provided by the traditional cable control4,5.
The information transfer capability of the cutaneous phantom was extended to a two-dimensional psychophysical display over a triangular area (using three stimulators) on the abdomen. The study optimized parameters and evaluated display resolution and dynamics in the context of the feasibility of providing range and azimuth information to a blind traveler6,7,8.
Massachusetts Institute of Technology, Cambridge, Massachusetts
1. Rothchild, R. D. and Mann, R. W., "An EMG Controlled, Proportional Rate, Force Sensing Elbow Prosthesis," Proceedings of the Symposium on Biomedical Engineering, Milwaukee, Wisconsin, June, 1966.
2. Alles, D. S., "A Method of Providing Kinesthetic Feedback from Prosthesis," 1st Canadian Medical and Biological Engineering Conference, September, 1966, Ottawa.
3. Alles, D. S., "Information Transmission by Phantom Sensations," IEEE Transactions on Man-Machine Systems, MMS-11, No. 1, March, 1970.
4. Mann, R. W., "Efferent and Afferent Control of an Electromyographic Proportional-Rate, Force Sensing Artificial Elbow with Cutaneous Display of Joint Angle," The Basic Problems of Artificial Prehension, Movement and Control of Artificial Limbs, London, 1968.
5. Mann, R. W. and Reimers, S. D., "Kinesthetic Sensing for the EMG Controlled 'Boston Arm,' " Proceedings of the Third International Symposium on External Control of Human Extremities, Dubrovnik, Yugoslavia, August, 1969; also IEEE Transactions on Man-Machine Systems, Vol MMS-11, No. 1, March, 1970.
6. McEntire. R. H., "Information Rate Via Vibro-Tactile Two Dimensional 'Phantom Sensation,' " Thesis (Ph.D.), Dept. of Mech. Engr., Mass. Inst. of Tech., June. 1971.
7. Mann. R. W., "Force and Position Proprioception for Prostheses," Control of Upper Extremity Orthoses and Prostheses (eds. Herberts, P., et al) to be published by Thomas Co., Ft. Lauderdale. Florida, 1973, (Presented at the International Symposium on the Control of Upper-Extremity Orthoses and Prostheses. Goteborg, Sweden, October 6 8. 1971).
8. Mann. R. W., "Prostheses Control and Feedback Via Noninvasive Skin and Invasive Peripheral Nerve Techniques," Houston Neurological Symposium, March 1-3, 1973.