Flexible Epidermal Electrodes for Intuitive Control of Powered Arm and Leg Prostheses
Abstract
Individuals with arm amputations typically use electrical signals (called EMG signals) to control motorized arm prostheses (also called myoelectric prostheses). EMG signals are generated when muscles contract. For prosthesis control, EMG signals generated by contraction of muscles on the residual limb are recorded by electrodes placed over the muscle, on the skin surface. Newer control systems rely on computer algorithms that can recognize complex patterns of EMG signals from many different muscles. This pattern recognition control technique allows for more natural control of more prosthesis movements. Over the last several years, we have developed advanced pattern recognition algorithms that provide excellent control of myoelectric arm prostheses and are now commercially available. Several powered leg prostheses are being developed by us and others. These devices are usually controlled by information from mechanical sensors built into the prosthesis, but such control systems do not always make the prosthesis do what the user wants it to, which can result in falls and injuries. We have developed algorithms that can combine EMG signal patterns from residual leg muscles with mechanical sensor data to provide excellent, reliable control of powered legs. This type of control system also feels very natural to use. Clinical use of these powerful pattern recognition control algorithms is hindered by difficulties in obtaining high-quality, reliable EMG signals. EMG electrodes typically consist of hard metal domes attached to rigid electronics, so they are often uncomfortable to wear. Also, if the electrode moves during prosthesis use, or comes away from the skin surface, EMG signals may change or be lost, which reduces the accuracy and reliability of the control algorithm. Although pattern recognition systems allow the user to quickly recalibrate the control system, needing to recalibrate frequently interrupts task performance and is frustrating. We have developed very thin elastic materials that can incorporate electrodes and other electronic components and that stick to the skin so that they move when then skin moves, almost like a tattoo. This means that the electrode-skin contact is always maintained and the electrode is always in the correct place, resulting in more consistent EMG signals. These thin devices (called Epidermal Electrode Systems, or EES) could be used to control both upper and lower limb prostheses; however, currently they require wires to transmit the EMG signals to the prosthesis. These wires frequently break, so they are not suitable for everyday use. Also, an EES requires power to record and process EMG signals and send the control information to the prosthesis. Our objective is to develop an EES that can generate its own power to record and transmit EMG signal data to the prosthesis using wireless technology. The rationale for this project is that improved EMG signal quality would significantly improve control of both powered arm and leg prostheses. An EES would work with any type of myoelectric arm prosthesis or powered leg prosthesis available now or in the future. Our project seeks to improve the control interface, i.e., the way that prosthesis users interact with the control system of their prosthesis, to enhance prosthetic function. Thus, this work addresses the Focus Area of Prosthetic and Orthotic Function. We expect that the benefits of a wireless EES, compared to a conventional EMG system, will be more reliable, consistent EMG signals; improved control of powered arm and leg prostheses; and enhanced user comfort. The potential clinical applications, in addition to control of upper and lower limb prostheses, include control of powered orthoses. An EES poses little or no risk to the user as it generates very little power, and the risk of an adverse reaction to the materials in the device is also very low. By the end of this study, we expect to have complete EES s
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Oct 29, 2018
- Source ID
- W81XWH1810668
Entities
People
- Levi J. Hargrove
Organizations
- Shirley Ryan AbilityLab
- United States Army