THE EFFECTS OF LIMB POSITION AND APPLIED LOAD ON HAND GESTURE CLASSIFICATION ACCURACY USING ELECTROMYOGRAPHY AND FORCE MYOGRAPHY
DOI:
https://doi.org/10.57922/mec.2489Abstract
Modern mechatronic upper limb prostheses are controlled using surface electromyography sensors (EMG) that are typically embedded in the prosthetic socket. However, when the user moves their device in space or interacts with an object, changes in electrode contact pressure can occur that work to the detriment of consistent and effective prosthesis control. Yet, we suggest that these pressure changes offer unique information that can be captured using force myography (FMG) and decoded to help classify intended prosthesis movements. Thus, the goal of this work was to investigate the feasibility of combining FMG with EMG to classify hand grasping movements in an able-bodied cohort and compare this combination to EMG and FMG alone. We hypothesized that FMG will capture complimentary information to the EMG data and when combined, will produce more robust classification accuracies when the user's limb moves in space or grasps objects of varying loads. We used a custom EMG+FMG armband and instructed N=21 participants to grasp objects of different weights at a variety of different positions using 4 different hand grasp movements. The results demonstrated that the average classification accuracy of EMG+FMG was statistically different and of higher classification accuracy when compared to EMG and FMG. It was also found that position and load affect classification accuracy together suggesting that control techniques that adapt to these changes are likely to produce more effective prosthetic control performance.
Downloads
Published
2024-08-15
How to Cite
[1]
P. R. Young, E. J. Winslow, G. K. Sagastume, M. A. Battraw, R. S. Whittle, and J. S. Schofield, “THE EFFECTS OF LIMB POSITION AND APPLIED LOAD ON HAND GESTURE CLASSIFICATION ACCURACY USING ELECTROMYOGRAPHY AND FORCE MYOGRAPHY”, MEC Symposium, Aug. 2024.
Conference Proceedings Volume
Section
Other
