Effects Of EEG - Based Exercise Training On Brain Function Network

The brain is a complex and dynamic system, many functions of which are located in the organizational structure of the brain. Affected by learning, training and experience and other factors; its structure and function will appear dynamic repair and recombination. Sports training as a specialized organization of the educational process, cooperate with the moral, intellectual, aesthetic, labor and technical education, is an indispensable component in the whole education. With the continuous development of various disciplines to integrate technology, researchers have found that long-term domestic and brain exercise training on human EEG activity will have particular impact. Studies of brain function network is a new way of in-depth understanding of brain function, and electroencephalogram (EEG) has been widely used in the study of brain functional network due to the high temporal resolution, can be monitored in real time, and easily accessible.This paper studies the 19-channel EEG signals of different levels of exercise training subjects. EEG was acquired in motor imagery task, and specific frequencies were extracted by the wavelet decomposition to calculate the value of each connection between EEG lead. Firstly, functional connection networks ofsport training group and control group were constructed using the Pearson correlation coefficient. Secondly, effective connection networks of sport training group and control groupwere constructed by the use of sparse Bayesian network. Finally the brain network measurements of the two groups were extracted and further study of the statistical properties of complex networks was performed, obtaininggraphs of each measurement changed in different thresholds.The study found that EEG functional connection networks and effective connection networksof different levels of sport training subjects in the beta frequency band exhibited more significant difference. Measurement analysis of functional connectivity network found thatconnectivity features of each node from two groups of subjects in four bands are not consistent; the average clustering coefficientin the alpha band is quite different, and nodes in functionalconnected network of training group are more likely to form groups; characteristic path length and modular index of two groups of subjects are no significant difference; betweenness at different frequencies are not the same. Measurement analysis of effective connectivity network found that sport training has a slightly larger degree of influence in the measurement of input degree, but the performance of different frequency bands are not the same; The influence of sport training on average clustering coefficient in beta and delta frequency bands has a certain performance, but the performance is unstable; characteristic path length and modular index of two groups in effective connection network are no significant difference; betweenness of two groups in effective connection network at different frequencies are consistent.The results show that the network construction methods used in the paper are able to describe the different brain functional states, becoming a new and effective method in analyzing and understanding the impact of sport training on the brain. This finding not only can promote our understanding of the relationship between human brain and motion information processing, but can master the law of brain electrical activities affected by sport training, providing a scientific evaluation method and a pertinency testing reference for personnel training.