Arm-Intent-Detection with Bio-Signals in Virtual Reality

Abstract

Looking at a huge number of amputees experiencing neuropathic phantom pain limb, this thesis presents the VITA 2.0 platform, targeting phantom pain relief by implementing the classical mirror therapy in virtual reality (VR). Broadly speaking the work surveys myography data collection, interpretation towards arm intention detection and visualisation of sEMG data in a virtual world. It also presents a way of low-cost, occlusion invariant arm constellation tracking to control the pose of a virtual arm. Apart from developing a new and versatile platform, a set of user focused experiments was conducted demonstrating the influence of the arm constellation for a successful intention detection. Whilst targeting healthy participants at this stage, future endeavours are set to include experiments with trans-radial amputees.