Applying Granger Causality and Graph Theory to Analysis of EEG Connectivity Change Caused by VNS
Abstract number :
1.077
Submission category :
1. Translational Research: 1D. Devices, Technologies, Stem Cells
Year :
2016
Submission ID :
194380
Source :
www.aesnet.org
Presentation date :
12/3/2016 12:00:00 AM
Published date :
Nov 21, 2016, 18:00 PM
Authors :
Tsuyoshi Uchida, Kyoto University; Koichi Fujiwara, Kyoto University; Takao Inoue, Yamaguchi University School of Medicine; Yuichi Maruta, Yamaguchi University School of Medicine; Manabu Kano, Kyoto University; and Michiyasu Suzuki, Yamaguchi University S
Rationale: To understand the physiological mechanism by which vagus nerve stimulation (VNS) suppresses epileptic seizures, it is important to analyze the influence of VNS on scalp electroencephalogram (EEG). In addition, such analysis may contribute to the development of a new prediction method for evaluation of VNS effectiveness prior to the implantation. In the present work, we analyzed changes in EEG connectivity caused by VNS with Granger causality (GC) and the graph theory. Methods: This work compared changes in EEG connectivity with and without the stimulus by VNS. We collected EEG data of three drug-resistant epileptic patients whose profile is shown in Table 1. The data collection and analysis were approved by the ethics committee of the faculty of Medicine and Health Sciences, Yamaguchi University. In the present work, GC was used as a connectivity index between the EEG channels, which shows whether one time series is useful in estimating future values of another time series or not. When a directed graph based on the GC values is constructed, it expresses the neural connection. In addition, in-degree, out-degree, and the numbers of edges coming into and going out from a vertex of the graph were used to evaluate the information flow intensity between the EEG channels. Results: Figure 1 shows the analysis results. In-degree values distributions of patient A were asymmetric bilaterally during the stimulus. Patient B had high out-degree values around the top head areas during the stimulus. In patient C, high out-degree values were observed particularly at the epileptic focus during the stimulus. These results suggest that the stimulus encourages information flows from an epileptic focus to other brain areas. Conclusions: In this work, we analyzed EEG connectivity change caused by VNS with GC and the graph theory. Our analysis results indicate that VNS changes neural connectivity particularly at an epileptic focus. In future works, additional clinical data will be collected for further investigation. Funding: JSPS KANENHI 26870314
Translational Research