Frequency dependent network flexibility analysis in epileptic brain based on phase locking value and resilience test: Analysis of frequency dependent information integration based on complex network

Abstract

Frequency component is critical for the brain to execute cognitive function by way of and cooperation of electrical signals. Complex network could visualize the neural system quantitatively and objectively based on graph theory. In this paper, we would focus on the study of broadband electroencephalogram recordings and combine phase locking value with resilience test to uncover frequency dependent network flexibility in the epileptic brain network. Phase locking value is efficient in detecting phase relationships in narrow band EEG waves by incorporating wavelet transform. Resilience test plays a role in the evaluating network's fragility by eliminating single node and its links randomly as well as in order. These methods are then applied on EEG signals recorded from the brain of human beings with four kinds of epilepsy disease. Results demonstrated that hierarchical order of network characteristic metrics are different in distinctive types of epilepsy disease; besides, the network's resilience are frequency sensitive in these pathological brain networks. Frequency dependent information transition and integration could be uncovered by these tools. Further research should pay attention to the evolution principle of these frequency reliance brain network, thereby promoting underlying working mechanism of these EEG signals in the brain.