A-mode Ultrasound Driven Sensor Fusion for Hand Gesture Recognition
Peter Boyd, Honghai Liu
Abstract
Traditionally, Surface electromyography (sEMG) has been the predominant method of sensing muscle activity in order to control myoelectric prosthesis. While many prosthesis control schemes used simple direct control, an ever increasing focus has moved to pattern recognition based approaches which promise a greater degree of natural control such as to further improve an amputees quality of life. Although pattern recognition based approaches have shown great promise, they have innate limitations due to changes that may occur during long term use which prevent clinical acceptance. Due to these limitations, researchers have increasingly investigated alternative modalities to provide more robust control schemes. A particular modality that has seen increasing interest is ultrasound based sensing due to its capability to better understand deep tissue activity.Within this research, A-mode ultrasound based sensing is proposed not as a replacement for sEMG based sensing but instead to augment and drive sEMG based sensing during activities that may otherwise prove challenging to traditional sEMg based control schemes.