Design and Clinical Observation of a Multisensory Stimulation Soft Hand Rehabilitation System

This study presents a multisensory stimulation hand rehabilitation system designed for stroke patients. The system consists of a soft hand exoskeleton, virtual reality (VR) scenes and a functional near-infrared spectroscopy (fNIRS) device. The exoskeleton provides motion assistance to each finger on the affected side. In the VR scenes, rehabilitation training tasks that need to be completed in coordination with the exoskeleton are set up. The combination of VR and the exoskeleton provides an immersive rehabilitation training experience, resulting in comprehensive and sufficient activation of the affected brain motor areas. The fNIRS device monitors the neural activity of the patient’s brain during rehabilitation training, enabling real-time evaluation of the neural effect. A total of 25 stroke patients (chronicity: subacute; Brunnstrom hand stage: 2–4; age: 59.3 ± 7.1 years) were recruited to use the system to complete grasp training, piano playing, and block grabbing tasks. Their kinematic performance and cortical activity were recorded and statistically analyzed (paired t-test). Compared to without system assistance, the participants who used the system presented an average 87% increase in joint range of motion on the affected fingers (p < 0.01), and the average pinch force increased to 9.4 ± 2.1 N (p < 0.01). Under multisensory stimulation, the motor cortex, and prefrontal cortex of the patient’s brain presented increased blood oxygen levels (p < 0.05). In addition, the activation of the temporal lobe by the music scene was significant (p < 0.05). This study aimed to investigate the integration of interventions and assessments after stroke.