An Improved Visibility Graph Analysis of EEG Signals of Alzheimer Brain

摘要

Focused on the issue of the poor robustness to the noise of the visibility graph (VG) algorithm, the limited penetrable visibility graph (LPVG), as an improved visibility graph algorithm, was applied to investigate the alteration of electrical activity in the brain of Alzheimer's disease (AD) patients. According to the LPVG algorithm, electroencephalograph (EEG) signals from Alzheimer's disease (AD) patients and the normal control subjects were mapped into complex network, then the topological network characteristics were extracted, thus the distinction of the two groups could be compared. Simulation results demonstrate that the LPVG algorithm applied in this paper could be regarded as a kind of effective method to characterize the abnormality of the topological structure of single EEG signal of AD, whose network was abnormal, as reflected in the decreased small-world properties. The conclusion drawn in the paper would provide help to detect AD clinically and study AD pathologically.