Control System Adaptation to Improve Upper-Extremity Prosthetic Limb Wear Time
Abstract
Rationale and Objectives: Upper limb amputation is a significant cause of disability that drastically limits an individual’s functional capabilities. Even the most technologically advanced prosthetic arms fail to adequately restore the functional capabilities of the lost arm. Despite being commercially available for over 40 years, control of powered prostheses remained virtually unchanged until 2013, when the first commercial pattern recognition controller was released. By using the pattern of muscle activity in the user’s residual limb to control the prosthesis, this controller provides more natural, intuitive control of a powered prosthesis. Currently, the controller must be retrained with new data each time the user wants to improve prosthesis performance. However, preliminary studies have shown that if some of the previous training data is kept, the performance of the prosthesis may be improved and the prosthesis may require less frequent training. Although preliminary results have been promising, this approach has not been clinically tested. We propose to implement this new control strategy (called adaptation) on the commercial version of a pattern recognition prosthesis controller and test how the new strategy affects the frequency of prosthesis recalibration and the amount of time a prosthesis is worn. Potential Impact: If the adaptation approach proves successful, it could quickly become part of a commercial pattern recognition control system, thereby offering improved prosthesis control and a decreased need for recalibration. In addition, this proposed study would result in baseline prosthetic usage statistics including how much individuals use a prosthesis at home and how electromyographic (EMG) signals change over the course of several months. These data will allow future hypotheses and research projects to be developed to further improve technologies for upper limb amputees. Target Population: The population that will be helped by this research includes individuals with major upper limb amputation. In the United States, upper limb amputations at the wrist level or high effect over 1200 new individuals each year, and the estimated population of individuals with these injuries is over 40,000. In addition, over 3% of all battle-related injuries for US Service members involve traumatic amputations, with over 200 upper limb amputations for Service members occurring between 2001 and 2011. By improving prosthetic control technology, we hope to give prosthesis users the ability to perform daily activities with greater ease and less frustration, thereby improving their quality of life. Clinical Applications, Benefits, and Risks: The current commercial pattern recognition system is being used by approximately 200 individuals with transradial, transhumeral, and shoulder disarticulation amputations. The proposed project would improve the performance of this controller without introducing additional risks. The control improvement would be configurable, meaning that users could turn the additional functionality off if they did not find it to improve the performance of their prosthesis. Projected Timeline: The timeline of this study is 3 years. At the end of the study period, we anticipate that a fully functional pattern recognition controller with adaptive training capabilities will be completed and submitted for U.S. Food and Drug Administration (FDA) approval. Once approved, the new version of the controller could be released to the commercial market immediately. The valuable prosthesis usage information and EMG signal data collected over the year-long study period could be made available to interested researchers once the study is completed. The benefits derived from the collection of these data would likely take several additional years to be appreciable. Benefit to Service Members and Veterans: The same benefits available to the general population of individuals with amputation would b
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Oct 29, 2018
- Source ID
- W81XWH1710645
Entities
People
- Blair Lock
Organizations
- Coapt (United States)
- United States Army