Investigation of robot assisted sensorimotor upper limb rehabilitation after stroke

Abstract

Strokes lead to both motor and sensory impairments in the neural circuit. Traditional stroke rehabilitation mainly focuses on motor restoration, but sensory participations together with motor recovery are frequently overlooked and poorly understood. Tactile perception is highly involved in the motor relearning process after a stroke, whereas tactile impairments and their contributions to the rehabilitative effects have received limited attention. Robots have been adopted for motor rehabilitation with high intensities. Robots have also been integrated with neuromuscular electrical stimulation (NMES) for effective motor relearning in our previous works. The objectives of this study include investigations on 1) the rehabilitation effectiveness when robot-assisted upper limb rehabilitation was integrated with enriched tactile sensory inputs, 2) the rehabilitation effectiveness when robot-assisted upper limb rehabilitation was integrated with sensory inputs induced by NMES, and 3) the extent of tactile impairments in the upper limb during textile fabric stimulation in stroke survivors. The study was divided into the following three parts: In the first part, we investigated the rehabilitation effects of the robot-assisted upper limb rehabilitation integrated with enriched tactile sensory inputs. Thirty-two participants suffering from chronic stroke received robot-assisted training either in the clinical service setting (n=l6) with an enriched rehabilitation environment, or in the well-controlled research setting (n=16). The results indicated that the functional improvements following the robotic hand training were comparable for the two groups, whereas the integration of enriched tactile sensory inputs led to greater independence in daily living and a more effective release in muscle tones. In the second part, we investigated the rehabilitation effects of the robot-assisted upper limb rehabilitation integrated with NMES. Thirty chronic stroke patients were randomly assigned to receive upper limb training with either an NMES robotic hand (n=15) or a pure robotic hand (n=l5). The results indicated that more effective distal rehabilitation could be obtained by the NMES robot than the pure robot, especially in the areas of lowered muscle spasticity and enhanced voluntary motor recovery and muscle coordination. In the third part, we investigated the extent of tactile impairments in the upper limb during textile fabric stimulation via electroencephalography (EEG) in stroke survivors. Twelve chronic stroke patients and fifteen healthy adults received 64-channel EEG detection with three different fabric stimuli on both sides of the volar forearms. The results supported the feasibility of using EEG to investigate tactile impairments following a stroke. The findings also suggested that the tactile impainnents after stroke could be represented by a shifted power spectrum, increased power intensity, and remapped sensory cortical areas. In conclusion, integrating the tactile sensory inputs into the robot-assisted training by providing enriched tactile sensory inputs and NMES could contribute to more functional recovery in the entire upper limb compared to robot-assisted training without tactile sensory integrations for chronic stroke. Moreover, EEG is capable of neurologically evaluating the extent of tactile impairments in stroke patients' upper limbs.