A0452
Title: Statistical network analysis for epilepsy MEG Data
Authors: Haeji Lee - Duksung women's university (Korea, South)
Jaehee Kim - Duksung Womens University (Korea, South) [presenting]
Sunhan Shin - Duksung Women's University (Korea, South)
Abstract: Network analysis is useful in understanding the structural and functional relationships between nodes by analyzing a mathematical model that represents a structure made up of a set of nodes and their connection forms. In neuroscience, network analysis using brain imaging data is actively being researched, but statistical methods for this are lacking. We intend to analyze statistical network analysis for epilepsy MEG data. The R program was used for the analysis, and the network was generated by using correlation coefficients of brain magnetic field signal data measured by 72 sensors. MEG data were collected from 44 Korean patients with epilepsy and 46 healthy controls. We compared the networks of healthy and patients including hub nodes, betweenness centrality, degree and network visualization. Time-varying networks were fitted using the TERGM(temporal exponential random graph model). The hub node in the patient group was a right posterior cingulate gyri node, while the hub node in the healthy group was a left anterior cingulate and paracingulate gyri node, both of which belong to the limbic system.
