Improving Cognitive Function By Neurofeedback Training In Mild Cognitive Impairment

Mild cognitive impairment(MCI)is a state between normal aging of brain and Alzheimer’s disease(AD).Its ad conversion rate is 10 times higher than that of normal people.Different from the irreversible characteristics of AD,it is a brain cognitive function disease that can delay the development.Therefore,it is urgent to find and intervene early.Neurofeedback training can improve individual cognitive function by regulating brain activity,which is a potential tool to achieve regulatory intervention and improve brain cognitive function.This paper aims to analyze the effect of neurofeedback training on the improvement of brain cognitive function in patients with MCI.Based on electroencephalogram(EEG)signal,the improvement of cognitive function of MCI patients after neurofeedback training was analyzed from two aspects of brain complexity and brain function connectivity,and the brain cognitive function evaluation system for MCI was completed.The neurofeedback training uses mind-control game,and the feedback indexes include relaxation index from the relative power of α rhythm and concentration index from the relative power ratio of β/α.During the training,the system will adjust the target characteristics according to the feedback index to improve the cognitive function of the subjects.In order to solve the problem of information loss of EEG complexity details,a continuous complexity and continuous sample entropy algorithm are proposed to analyze the influence of neurofeedback training on the complexity of brain.Continuous complexity and continuous sample entropy are achieved by segmenting EEG signals and calculating the eigenvalues of each segment respectively.The results show that after the neurofeedback training,the continuous complexity mean and the continuous sample entropy mean of EEG signal are significantly increased(P < 0.05).The continuous complexity and the continuous sample entropy characteristic curve are significantly higher than that before the training,and the curve fluctuation amplitude is reduced,which indicates that neurofeedback training can improve the complexity of brain functional state,and the effect is stable.Based on the coherence and phase synchronization,the coupling and synchronization of EEG signals among the electrode pairs of δ,θ,α and β rhythms is analyzed.The eigenvalue matrix and brain function connection diagram are obtained,and the effect of neurofeedback training on brain function connection is studied.After neurofeedback training,the characteristic values of EEG signal of each rhythm are increased,but only the characteristic values of β rhythm keep increasing(coherence: 0.377?0.138,0.422?0.124,0.433?0.139,P=0.003<0.05,phase synchronization index: 0.415?0.154,0.451?0.138,0.455?0.150,P=0.075),and the number of electrode pairs with significant difference before and after training is the most(n=7),indicating that the improvement effect of β rhythm brain function connectivity is better.It can be seen from the functional connection diagram that the functional connectivity within and between hemispheres has improved after training,especially in the frontal region.The analysis of coherence and phase synchronization can objectively reflect the brain functional connectivity.Combined with the results of EEG complexity analysis,neurofeedback training has a positive effect on improving cognitive function of patients with MCI.Based on the graphical user interface(GUI)of MATLAB,the design and implementation of brain cognitive function evaluation system for MCI are completed.The system is mainly divided into four parts: data reading,preprocessing,complexity analysis and functional connectivity analysis.It can display the results of brain cognitive function analysis and provide a tool for evaluating the complexity and functional connectivity of brain.