Currently upper-limb amputees can only operate a single degree-of-freedom at a time with myoelectric prostheses. This is very inadequate, especially for high-levels of amputation such as shoulder disarticulation(SD) where multiple functions need to be controlled. We postulate that the residual brachial plexus nerves in a SD amputee can be grafted onto separate regions of the pectoralis major (pmajor) muscle and that these nerve-muscle grafts could provide additional myoelectric control signals that are physiologically related to the functions they would be controlling in the prosthesis. This would allow simultaneous control of multiple degrees-of-freedom with a more natural feel. The technique has great potential for improving the control of myoelectric SD prostheses. The key to success with this technique will be the ability to record independent surface EMG signals from each of the nerve-muscle grafts. In order to study EMG signal independence in the chest, a series of finite element (FE) computer models of EMG signal propagation in the chest will be developed and validated with experimental data. Using FE analysis, it is possible to simulate surface EMG signals under a range of different conditions. Effects such as muscle anatomy, biological tissue properties and recording electrode configuration will be investigated in a manner not possible using experimental methods. First, FE analysis will be used to investigate the relationship between surface EMG signal independence and the geometry of the active muscle, neighboring muscles and other tissues near the recording site. This will be accomplished with a series of generalized planar FE models. Next, finite element analysis will be used to determine the effect of anatomical manipulations for improving surface EMG signal independence including removal of fat, concentrating muscle tissue at recording sites and insulating muscles with a layer of fat. Finally, the subject-specific models will be used to simulate the nerve-muscle graft technique and test the feasibility of this novel approach. Anatomical manipulations to enhance surface EMG signal independence will also be tested with the subject-specific models.
Showing the most recent 10 out of 28 publications
