Feedback Control to a Static Target Angle in the Middle Finger Metacarpophalangeal Joint Using Functional Electrical Stimulation. Issue 15 (13th September 2020)

• Record Type: Journal Article
• Authors: Watanabe, Kyosuke; Oka, Makoto; Mori, Hirohiko
• Is Part Of: International journal of human-computer interaction. Volume 36:Issue 15(2020)
• Page Start: 1476
• Page End: 1486
• Publication Date: 2020-09-13
• Subjects: Human-computer interaction -- Periodicals; 004.01905
• DOI: 10.1080/10447318.2020.1758879
• Languages: English
• ISSNs: 1044-7318

Abstract

ABSTRACT: In this study, we proposed a feedback control method for a finger joint angle, which has been considered difficult, via functional electrical stimulation using the surface electrode method. If detailed finger movement can be realized by functional electrical stimulation using the surface electrode method, this can be applied in various situations such as musical instrument performance, daily movements, and so on. The control method was designed based on the following three proposals. First, the non-linear relationship between the electrical stimulation intensity and joint angle change was absorbed by constructing polygonal line regression models of the relationship between the electrical stimulation intensity reaching each muscle and the flexion and extension force of the middle finger's metacarpophalangeal joint. Second, the concepts of the muscle antagonist ratio and muscle antagonist sum were adopted to translate the input/output relationship to a single-input/single-output system while maintaining joint stiffness. Finally, the proportional–integral–derivative control parameters were determined using the Ziegler–Nichols ultimate sensitivity method. Using the control method based on the aforementioned three proposals, we conducted feedback control experiments for three static target angles in the middle finger's metacarpophalangeal joint. Results show that at all target angles, greater than 80% of the participants were able to control the joint with nearly no error. Therefore, the control method proposed in this study can be used to control an arbitrary static target joint angle with high accuracy via functional electrical stimulation.