Grasp and release with surface functional electrical stimulation using a Model Predictive Control approach.

Stroke often has a disabling effect on the ability to use the hand in a functional manner. Accurate finger and thumb positioning is necessary for many activities of daily living. In the current study, the feasibility of novel FES based approaches for positioning the thumb and fingers for grasp and release of differently sized objects is evaluated. Assistance based on these approaches may be used in rehabilitation of grasp and release after stroke. A model predictive controller (MPC) is compared with a proportional (P) feedback controller. Both methods are compared on their performance in tracking reference trajectories and in the capability of grasping, holding and releasing objects. Both methods are able to selectively activate the fingers such that differently sized objects, selected from the Action Research Arm test, can be grasped. The MPC method is easier to use in practice, as this method is based on a single identification of a model of the biological system. The P-controller has more parameters which need to be set correctly, and therefore needs more time to initialise. The current results are very promising. Evaluation in patients will be done to explore the possibilities to apply these methods in rehabilitation of grasp and release after stroke.