Deep and Surface Sensor Modalities for Myo-intent Detection

Abstract

Electromyography is the gold-standard among sensors for prosthetic control. However, stable and reliable myocontrol remains an unsolved problem in the community. Amid improvements currently under investigation, one focuses on alternative or complementary sensors. In this study, we compare different techniques, recording surface and deep muscle activity. Ten subjects were involved in an experiment in which three different modalities were attached on their forearm: force myography, electro-impedance tomography and ultrasound. They were asked to perform wrist and grasp movements. For the first time, we evaluate and compare in an offline analysis these three different modalities while recording several hand gestures.