Nye kontrolmetoder for neurale proteser til overarm / New control methods for upper-arm neural prostheses

In rehabilitation of humans with motor impairment the aim is to activate joints in a controlled way so as to restore as much motor function as possible. The control strategies applied in most rehabilitation devices so far have been developed largely in relation to the design of machines rather than to the design of sensory motor systems. Some recent rehabilitation results show that integrating more natural and timely interventions could lead to higher efficiency. It raises questions: How general were these findings? In assistive devices, should we always struggle to mimic the relevant natural control system on the assumption that it has been optimized in the course of evolution? Could we always mimic nature, given that our abilities to reproduce components of the neuromuscular system are limited? This project integrates the knowledge from biological control methods with the most advanced computer techniques, modeling, and simulation in order to design a rather simple, yet possibly practical assistive system. The model that we follow has the following characteristics: 1) the control system has a hierarchical structure; 2) feedback loops connect the lower levels with the higher ones in order to tune the descending (efferent) commands; 3) time delays in the feedback loops require combining feedback and predictive, open loop modes of control; and 4) the number of degrees of freedom in a motor system is excessive; consequently, the process of control can be regarded as overcoming the ambiguity caused by redundancy. The project will specifically approach the development of new assessment methods for determining the rationale why some techniques contribute more than others to the recovery of function. This will include brain mapping (EEG and MEG), and determining connectivism between biomechanical and electrophysiological variables. Funded by the Danish National Research Foundation. (Mirjana Popovic, Dejan Popovic, Thomas Sinkjær)