Passive reach and grasp with functional electrical stimulation and robotic arm support

Rehabilitation of arm and hand function is crucial to increase functional independence of stroke subjects. Here, we investigate the technical feasibility of an integrated training system combining robotics and functional electrical stimulation (FES) to support reach and grasp during functional manipulation of objects. To support grasp and release, FES controlled the thumb and fingers using Model Predictive Control (MPC), while a novel 3D robotic manipulator provided reach support. The system's performance was assessed in both stroke and blindfolded healthy subjects, where the subject's passive arm and hand made functional reach, grasp, move and release movements while manipulating objects. The success rate of complete grasp, move and release tasks with different objects ranged from 33% to 87% in healthy subjects. In severe chronic stroke subjects especially the hand opening had a low success rate (<;25%) and no complete movements could be made. We demonstrated that our developed integrated training system can move the passive arm and hand for functional pick and place movements. In the current setup, the positioning accuracy of the robot with respect to the object position was critical for the overall performance. The use of a higher virtual stiffness and including feedback of object position in the robot control would likely improve the relative position accuracy. The system has potential for post-stroke rehabilitation, where support could be reduced based on patient performance which is needed to aid motor relearning of reach, grasp and release.