P 68 Novel control concepts and motor re-learning strategy in neurorehabilitation – practically-oriented approach

Abstract

Background and motivation Persons suffering from functional impairment, due to cerebral palsy, stroke, or Parkinson’s, often have not reached their full potential for recovery which often is a reason for injures and loss of life due to fall. Motor skill learning and retention of motor skills can be enhanced if a patient assumes control over practice conditions, e.g. timing of exercise instructions and feedback. In our study, we follow a novel conceptual framework ( Despotova and Kiriazov, 2015 ) for optimal control learning of goal-directed motion tasks, like reaching, standing up and walking. Currently we are designing an adequate control strategy for posture stabilization tasks, too. ( Despotova et al., xxxx ). Generally speaking our research on neurorehabilitation is based on underlying principles and novel concepts of neurophysiology, multi-segmental dynamics, optimal control theory, and control learning (CL). Methods Our approach is based on the real dynamics and its control concepts are biologically plausible. Moreover, the proposed CL approach has the necessary mathematical guarantees for its feasibility and convergence. For dynamic goal-directed movements, we can apply an efficient CL scheme which has the following main stages: • Choose a set of appropriate test control functions • Define the most relevant pairs of control parameters and controlled outputs • Perform control parameter optimization in training exercises Of primary importance is to define a set of variables that best characterize the dynamic performance in the required motion task. We make corrections in the control parameters until reach the target, applying efficient, natural learning algorithms. Results Our control concepts have been verified in numerous computer simulations. The proposed approach is successfully applied using human body models with 2, 3, 4 and 6 degrees of freedom representing reaching ( Fig. 1 ), standing up, and walking motion in sagittal plane, Fig. 2 . In Fig. 1 , the convergence to the target in a reaching motion task is achieved after a very small number of trials. The efficiency of our CL algorithm is confirmed also in a real body experiment – healthy subject performing fast reaching (ballistic-like) movements. Special attention in our current research is devoted to the challenging problem of CL in the real, three-dimensional human locomotion. We can perform proper decomposition of this complex motion task into several goal-directed movements and apply the proposed CL scheme for each of them. Discussion We believe that the proposed approach can be used to rebuild the so-called internal models (cortical reorganization) by proper training procedures. The work outlined here can provide a fundamental understanding of motor learning and may lead to the development of optimal strategies for efficient neuro-muscular rehabilitation. In addition we have found that our approach is very promising and could be efficiently applied to the more complex, human/robot systems. Various assistive devices may be used for restoration of human movement functionality, such as passive and active orthoses, robotic exoskeletons, EMG, FES, brain-computer interfaces (BCI).