Simultaneous, Proportional and Independent sEMG-Based Hand-Wrist Prosthesis Control

Period of Performance: 09/04/2016 - 08/31/2017

$742K

Phase 2 STTR

Recipient Firm

Liberating Technologies, Inc.
HOLLISTON, MA 01746
Principal Investigator

Abstract

? DESCRIPTION (provided by applicant): Surface electromyogram (sEMG) controlled powered hand/wrist prostheses are used by upper limb amputees to return partial upper limb function. Conventional transradial prostheses can use surface EMG amplitudes from the residual forearm flexors and extensors to control hand closing and opening, respectively. Additional degrees of freedom (DoFs), e.g., wrist rotation, are not controlled simultaneously in conventional commercial systems. Rather, prostheses apply EMG-based or mechanical mode switching, so that the same EMG sites sequentially control the additional functions. This limitation is problematic, as many basic tasks- e.g., drinking from a cup, combing hair-require the simultaneous activation and control of two or more joints. It was reported that the ability to control ... coordinated motions of two joints at the same time was the second highest priority improvement desired by these users. The first priority was for a rotating wrist-a necessary prerequisite for two DoF control, and a device that is already commercially available. A number of research studies have reported on the development of proportional, simultaneous and independent 2-DoF controllers, initially using upwards of 64 or more high-density electrodes (electrodes that were not intended for clinical use) and recently using as few as 7-8 commercial electrodes. For practical clinical use, however, fewer electrodes are necessary and a method is needed to determine where to locate them within the socket. The manual palpation and trial-and-error methods used to locate two myoelectrodes for existing 1-DoF systems is inadequate for 2-DoF electrode site selection. In our Phase 1 work, we demonstrated the feasibility of a system for electrode site selection that controls two DoFs with only four commercial electrodes. A calibration procedure using a 16-channel EMG system without the prosthetic socket is used in the laboratory (or, in the prosthetist's office in commercial implementation) to select EMG sites and relate their electrical activity to 2 DoFs. In this Phase 2 application, we propose to transitin this method to a commercial product. Aim 1 will develop the necessary hardware/software to prototype the embedded commercial controller as well as the clinical fitting/training system. Aim 2 will refine the site selection and control algorithms. Our Phase 2 work will focus on utilizing only the affected limb (for a more direct control relationship) and on determining ways to rapidly calibrate a 2-DoF controller (in the laboratory and in the field). Aim 3 will evaluate the prototyp devices and algorithms developed in Aim 1 and 2 in a laboratory study and then in a pilot field study, facilitating commercialization.