|
| Upper-Limb Prosthetics | |||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||
| Implantable Myoelectric Sensors (IMES) for Prosthesis Control Richard F. ff. Weir, PhD, Principal Investigator Co-Investigators: Funded by: National Institutes of Health (NIBIB) Purpose We have received funding for a 5 Year Bioengineering Research Partnership [BRP: 1 R01 EB01672-01, "Multifunction Prosthesis Control using Implanted Sensors"] from the National Institutes of Health (NIBIB) to develop a multi-channel/ multifunction prosthetic hand/arm controller system capable of receiving and processing signals from up to sixteen "BIONR-like" Implanted MyoElectric Sensors (IMES). Background Muscles when they contract generate an electric signal (the electromyogram or EMG) as a by-product of their contraction. This signal can be detected and measured using appropriate electronics and used to control a prosthesis (called "myoelectric prosthesis control"). Traditionally, surface electrodes are used to detect these EMG signals. However these signals are prone to cross-talk and it hard to distinguish the contribution to the measured signal from the various different muscles of the forearm. Currently, surface EMG on the forearm will only provide up to four independent channels for control due to crosstalk, but there are eighteen extrinsic muscles of the forearm involved in the control of the hand and wrist. Methodology Using implanted myoelectric sensors (RFB BIONs, see figure above), electromyographic (EMG) signals from multiple muscles can be sensed simultaneously and telemetered out to a controller in the prosthesis to provide multi-function prosthesis control. The implantable sensors will be individually addressable, hermetically encapsulated, leadless electrical devices that will be small enough to be injected percutaneously into muscles (2 mm diameter by 15 mm long). They will receive their power, digital addressing, and command signals via forward telemetry from an external transmitter coil laminated into the prosthetic socket. The implants will send their EMG data to the prosthesis controller via reverse telemetry.
The ability to sense EMG from multiple muscles raises issues as to the best way to use this information for control. We are using fine wire intra-muscular recordings in human subjects in conjunction with reach-to-grasp experiments to study the most appropriate method for control, be it one muscle one function, one pattern of muscle activation-one function, or possibly the use of combinations of postural synergies. This is an ambitious project that involves collaborations with Illinois Institute of Technology to develop integrated circuits for the sensors, a company called Sigenics Inc. to develop the external reader coil, and the Rehabilitation Institute of Chicago to help with both human subject and animal experiments. | ||||||||||||||||||||||||||||
| ||||||||
