Finger Displacement Sensing: FEM Simulation and Model Prediction of a Three-Layer Electrode Design.

There is a growing and significant interest in home-based therapy or telerehabilitation for physiological disabilities, for example, as a result of stroke. These technologies allow more flexibility in implementing rehabilitation sessions and offer the potential to reduce the economic burden of outpatient appointments and reduce the reliance on healthcare systems. However, extra effort needs to be made to make such systems effective. This paper investigates the feasibility of a home-based device, which is capable of detecting minute movements of patients’ fingers in regular training and testing sessions, addressing ease of use, motivation of practice, as well as feedback and guidance on performance. Toward this aim, the measuring techniques that are compatible with these targets were investigated. Based on a customizable three-layer electrodes design for use with an MGC3030 motion sensor IC, a finite-element method simulation in COMSOL Multiphysics and a nonlinear regression analysis using MATLAB were carried out. Four nonlinear equations were introduced to describe the motion of the index and middle fingers in the electrical field (E-field) generated. The form of the prediction models agrees with the hypothesis based on the quasi-static E-field sensing theory. In addition, these prediction models fit well with the relationship between finger distance and the voltage signals detected. With the prediction model, the targeted system is capable of detecting combined movements of two fingers at a resolution of 0.94 mm in a portable smart device for robust hand rehabilitation. [ABSTRACT FROM AUTHOR]