Research On Intelligent Stroke Upper Limb Rehabilitation System Based On Brain Computer Interface

Brain-computer interface(BCI)is a communication mode that acquires the electrical signal of human brain and identifies the intention,constracts the direct interaction between human brain and external equipment.BCI has been widely used in the fields of rehabilitation medicine,game entertainment,special work and so on.This paper combines the BCI system with the rehabilitation treatment of upper limb dyskinesia of stroke patients,patients can control the stroke treatment equipment by themselves,accelerate the repair speed of damaged neurons and improve the quality of rehabilitation treatment.The main contents of this paper include:1)Aiming at the problems of real-time EEG signal acquisition and equipment portability in the use of BCI system,we research the EEG signal acquisition technology and independently develops 16-channel digital EEG acquisition machine JSAI-16.The device can acquire real-time electrical signal from 16 electrode points in the human brain at a sampling rate of 250 Hz.2)We research the evocation principle of motor imagery paradigm,and design the workflow of intelligent stroke upper limb rehabilitation system according to the stroke rehabilitation characteristics,including four processes: preparation state,imagination state,dynamic state and rest state.According to the characteristics of motor imagery signals,we propose a classification framework of motor imagery state recognition based on brain network analysis and multilayer convolutional neural network.In order to verify the classification performance of the algorithm,we design an on-line EEG classification experiment including 10 subjects,the recognition accuracy of algorithm is 86%,shows excellent classification performance.3)Aiming at the problems of traditional treatment methods in the rehabilitation of upper limb movement disorders,we independently develope neuromuscular electrical stimulator JSAI-NMES and soft robot rehabilitation glove JSAI-SSRG as stroke rehabilitation equipment.Form an intelligent stroke rehabilitation BCI system with 16-channel digital EEG acquisition machine and motor imagery state recognition classification algorithm.The neuromuscular electrical stimulator can output the electrical stimulation signal with 5-gear frequency and stepless amplitude adjustment,treat the motor function of the patient’s arm;The soft robot rehabilitation glove has 9kinds of motion rehabilitation actions and 3 working operation modes,and carry out the rehabilitation treatment of stroke patients’ palm motor function to alleviate the symptoms of hand curling.Through the organic combination of various devices and EEG classification algorithms,we can effectively treat the upper limb dyskinesia of stroke patients and help them return to the daily life.