Multi-Day Real-time Myoelectric Control using Intramuscular EMG

Research output: Chapter in Book/Report/Conference proceedingMeeting abstractpeer-review

52 Downloads (Pure)

Abstract

Pattern recognition (PR) algorithms have been widely studied in the literature and have shown promising results in acute experiments. However, the clinical usability is of PR based myoelectric control is limited by unsatisfactory robustness to non-stationarities. Despite the body of evidence that time has an effect on classification performance, the influence of time on real-time performance on intramuscular EMG recordings has received less attention in the literature. The aim of this study was to investigate the real-time performance of intramuscular recordings over multiple days and to propose an experimental protocol to test three different training strategies using a Fitts’ law approach. Three pairs of wire electrode were inserted in the extensor and flexor muscles (Figure 1) and kept inside these muscles for five continuous days in two able-bodied subjects. Four Fitts’ law metrics (Table 1) were assessed using three train-test strategies: (i) An artificial neural network (ANN) classifier was trained on data collected from the previous day and tested on present day (BDT) (ii) Trained and tested on the same day (WDT) and (iii) trained on all previous days including present day and tested on present day (CDT) for five days. For both subjects, results are presented in Table1. The results of this study have confirmed the previous offline findings stipulating that daily calibration is important for stable performance. Nevertheless, results from two subjects cannot make the basis for a meaningful conclusion. We recommend that PR control scheme be tested and compared over several days to really capture the behavior of the system.
Original languageEnglish
Title of host publicationTrent Prosthetics – Trent International Prosthetics Symposium 2019
Place of PublicationSalford Manchester, UK
Publication statusPublished - 2019

Fingerprint

Dive into the research topics of 'Multi-Day Real-time Myoelectric Control using Intramuscular EMG'. Together they form a unique fingerprint.

Cite this