Sonomyographic Upper Limb Prosthetics: A New Paradigm

Overview

The vast majority of all trauma-related amputations in the United States involve the upper limbs. Approximately half of those individuals who receive a upper extremity myoelectric prosthesis eventually abandon use of the system, primarily because of their limited functionality. Thus, there continues to be a need for a significant improvement in prosthetic control strategies. The objective of this bioengineering research program is to develop and clinically evaluate a prototype prosthetic control system that uses imaging to sense residual muscle activity, rather than electromyography. This novel approach can better distinguish between different functional compartments in the forearm muscles, and provide robust control signals that are proportional to muscle activity. This improved sensing strategy has the potential to significantly improve functionality of upper extremity prostheses, and provide dexterous intuitive control that is a significant improvement over current state of the art noninvasive control methods. This interdisciplinary project brings together investigators at George Mason University, commercial partners at Infinite Biomedical Technologies as well as clinicians at MedStar National Rehabilitation Hospital. The investigators will optimize and implement algorithms for real-time classification and control with multiple degrees of freedom (DOF) using a miniaturized ultrasound system incorporated into a prosthetic socket. The investigators will then compare control performance between and sonomyography and myoelectric control (both direct control and pattern recognition) using a virtual environment as well as for performance of tasks related to activities of daily living. The investigators have two specific aims. Specific Aim 1: Compare between sonomyography and myoelectric direct control Specific Aim 2: Compare between sonomyography and pattern recognition with velocity control The successful completion of this project will lead to the first in human evaluation of an integrated prototype that uses low-power portable imaging sensors and real-time image analysis to sense residual muscle activity for prosthetic control. In the long term, the investigators anticipate that the improvements in functionality and intuitiveness of control will increase acceptance by amputees.

Specific Aim 1: Compare between sonomyography and myoelectric direct control The investigators would like to test the null hypothesis that there is no difference in performance on virtual environment and clinical outcome measures between sonomyographic and direct myoelectric control with mode-switching. Participants: The investigators will recruit a total of 10 unilateral trans-radial amputees based on the inclusion and exclusion criteria. These participants will be naïve to use of myoelectric prostheses. This study is focused on trans-radial amputees who represent the largest population of users. Protocol: After informed consent is obtained, the investigators will perform a baseline evaluation of the subject's functionality using the Orthotics and Prosthetics User Survey Upper Extremity Functional Status (OPUS-UEFS). Then subjects will be randomized to either the Experimental first group (sonomyography control) or Standard first group (direct control). Subjects will be tested using both control systems following a two-by-two AB/BA crossover design. The sonomyographic system will include a custom-fitted test socket incorporating ultrasound transducers and controlling a TASKA prosthetic hand. The direct control system will include a custom-fitted test socket incorporating dual-site electrodes and controlling a TASKA prosthetic hand. All socket fittings will be performed by a board-certified prosthetist as part of this study. After being fit with the sonomyography system, subjects in the Experimental first group will undergo 4 hours of standard occupational therapy training by a certified therapist. The training will be distributed over 2 days with 2 hours of training per day. At the end of each hour of training, subjects will perform a Fitts' Law task in a virtual environment. The task and the primary and secondary outcome measures are described below. After the training has been completed, subjects will evaluated using a series of primary and secondary clinical outcome measures, as described below. Finally, subjects will complete the Satisfaction with Prosthesis section of the Trinity Amputation and Prosthesis Scales - Revised (TAPES-R) to assess satisfaction with the sonomyography system. After a washout period of 4 weeks, subjects will cross-over to the Standard system. Subjects will be fitted with the direct control system and undergo the same training and testing procedure described above. Subjects in the Standard first group will be trained and tested with the direct control system first, and then cross over to the sonomyography system. Details of Fitts' Law task protocol: Subjects will move their limb into one of 4 positions following a display indicated on a screen (Fig. 16 in Research Strategy). Achieving each target position requires a different amount of shoulder abduction. When participants have reached the required limb position, an annulus-shaped target will be displayed. Subjects will use their prosthesis to perform a Fitts' Law task that involves controlling the position and size of a ring-shaped cursor in order to match the target. The Fitts' Law task will have one, two, or three degrees of freedom, where each degree of freedom is mapped to a specific motor intent. For example, wrist flexion can control horizontal cursor movement, pronation/supination can control vertical cursor movement, and hand open/close can control the cursor's radius. Subjects will be allowed to activate the degrees of freedom sequentially or simultaneously. The annulus width (W) and distance (D) from the initial cursor position will be varied to create six indices of difficulty according to ID = log2(1+ D/W). There will be three levels of target complexity based on the number of degrees of freedom that must be activated to achieve the task (i.e., one degree of freedom targets require activity in only one degree of freedom, etc.). Each possible combination of levels of difficulty (6) and target complexities (3) will be repeated 3 times per limb position. Subjects will have 10 seconds to match the target and hold for two seconds, or the attempt will be considered failed. After 12 seconds, the display will indicate the next target. Data Analysis: Data will be evaluated to determine the distribution and the variability of performance within and between the different prosthesis systems. The investigators will test whether the sonomyography and direct control systems are equivalent using the primary outcome measures. Secondary outcome measures will also be explored in order to develop a more detailed understanding of differences in performance between the systems. If significant differences are found in baseline functional measures using OPUS-UEFS, the investigators will perform stratified analyses on our primary and secondary outcome measures. Specific Aim 2. Compare between sonomyography and pattern recognition with velocity control Participants: The investigators will recruit a total of 6 unilateral trans-radial amputees based on the inclusion and exclusion criteria noted above. These participants will be experienced users of myoelectric pattern recognition control (IBT Sense). This study is focused on trans-radial amputees who represent the largest population of users. Protocol: After informed consent is obtained, the investigators will perform a baseline evaluation of the subject's functionality using the Orthotics and Prosthetics User Survey Upper Extremity Functional Status (OPUS-UEFS). Then subjects will be randomized to either the Experimental first group (sonomyography control) or Standard first group (pattern recognition control). Subjects will be tested using both control systems following a two-by-two AB/BA crossover design. The sonomyographic system will include a custom-fitted test socket incorporating ultrasound transducers and controlling a TASKA prosthetic hand. All socket fittings will be performed by a board-certified prosthetist as part of this study. The pattern recognition system will include the subject's clinically-prescribed socket and commercially-available pattern recognition technology (IBT Sense). The pattern recognition system will not be fitted as part of this study, as subjects will already be trained users prior to enrolling in the study. After being fit with the sonomyography system, subjects in the Experimental first group will undergo 4 hours of standard occupational therapy training by a certified therapist. The training will be distributed over 2 days with 2 hours of training per day. At the end of each hour of training, subjects will perform a Fitts' Law task in a virtual environment. The task and the primary and secondary outcome measures are described below. After the training has been completed, subjects will evaluated using a series of primary and secondary clinical outcome measures, as described below. Finally, subjects will complete the Satisfaction with Prosthesis section of the Trinity Amputation and Prosthesis Scales - Revised (TAPES-R) to assess satisfaction with the sonomyography system. After a washout period of 4 weeks, subjects will cross-over to the Standard system. Subjects will not undergo training, and will instead be immediately tested with the pattern recognition system using both the Fitts' Law task and clinical outcome measures. Subjects in the Standard first group will be tested with the pattern recognition system first, and then cross over to the sonomyography system. Data Analysis: Data will be evaluated to determine the distribution and the variability of performance within and between the different prosthesis systems. The investigators will test whether the sonomyography and pattern recognition control systems systems are equivalent using the primary outcome measures. Secondary outcome measures will also be explored in order to develop a more detailed understanding of differences in performance between the systems. If significant differences are found in baseline functional measures using OPUS-UEFS, the investigators will perform stratified analyses on our primary and secondary outcome measures.