The Key Mechanisms Of Neurofeedback Response In Brain Computer Interaction Based Motor Training And Its Application For Stroke Rehabilitation

Movement(or Motor)is one of the most important forms of human survival,labour and communication with the outside world.However,some accidents or diseases can lead to the loss of partial or total motor functions of the human beings.Specifically,stroke has become the first major disability disease in China and the whole world.Its young onset,high incidence of sequelae,long-term illness and huge medical expenditure make the society overwhelmed and difficult to prevent.Therefore,the researches of stroke rehabilitation are of great scientific and social value.However,the conventional rehabilitation therapies are difficult to induce the synchronous coupling of corticomuscular function,especially lack of the theoretical guidance of the mirror neuron system and neuroplasticity,resulting in the limitation of the final rehabilitation effectiveness.Recently,the brain-computer interface(BCI)and/or other emerging human-machine technologies based neurofeedback training(NFT)method has emerged to make it quantitatively observable for the information of central nervous system and real-time perceptible for limb movement,thereby promoting functional reconstruction of the whole neural pathway and motor system.It brings new hope for the development of motor neural rehabilitation technology.With urgent needs of stroke rehabilitation,this paper focuses on the scientific and technical problems of the regularity of neural induction and the mechanism of plasticity rehabilitation in motor NFT,as well as the efficient matching between neural information decoding and feedback regulation.The following research contents are specifically conducted: by the combination of EEG and NIRS based brain blood oxygen,we compared and probed the neural response mechanism of different feedback training modes to reveal the coupling regulation effect of BCI.And then,the BCI based visual-haptic motor NFT method were proposed and designed.We combined EEG-NIRS features to analyze the neural induction mechanism and the effectiveness of motor training of this method.The EEG-EMG responses of the new motor NFT methods under the influence of time-history factors and their coupling relationships were further developed.Finally,an integrated multi-modal motor NFT system was designed.And the typical application of long-term stroke rehabilitation was preliminarily attempted to verify its feasibility and effectiveness in clinical motor training.The main results show that BCI technology indeed plays a key role in motor NFT,and driving functional electrical stimulation(FES)could induce stronger electroencephalographic and cerebral hemodynamic activations.The BCI based NFT was proposed and implemented by combining with visual scene and haptic sensation of electrical stimulation.It achieved a significant improvement(～9%)in its classification performance.The cortical activations induced by motor imagery were also significantly enhanced.The combination of EEG-NIRS features could significantly optimize classification performance(reaching ～89%).Moreover,we found the multi-band coupling changes and regularity of EEG-EMG.Comparing with non-feedback training condition,the cortical-muscle activations of stroke patients were significantly improved after a period of the new motor NFT.The application of integrated multi-modal motor NFT system effectively improved the clinical behavior and motor neural function of typical stroke patients.In summary,we proposed and finally realized the "training and evaluation" integrated method of multi-modal motor NFT system.And we preliminarily achieved its validation of application effectiveness in stroke rehabilitation.These relevant results are expected to support the development and application of the new generation of neural rehabilitation technology.