Decoding 2D kinematics of human arm for body machine interfaces

Body-machine interface provides stroke and spinal cord injured patients a mean to participate in their activities of daily livings (ADLs). In this paper, electrophysiological signals from the human upper limb are used as a control interface between the user and a virtual robotic wheelchair. There is a general perception that these body signals contain an insufficient level of information for decoding or reconstructing kinematics of multi-joint limb activity. In this paper we present the results obtained in our virtual reality laboratory at Macquarie University, showing that non-invasive upper limb signals from high density wearable sensing shirt can be utilized to continuously decode the kinematics of 2D arm movements. Our results also show that body signals contain an information about the neural representation of movement. Moreover, they provide an alternative way for developing non-invasive body-machine interfaces, which have diverse clinical applications and access to these signals may provide understanding of functional brain states at various stages of development and aging