Epilepsy is one of the most common neurological disease, approximately 65 million people diagnosed with epilepsy in the world. For the removal of epileptogenic region in particular and epilepsy diagnosis in general, accurate localization of the epileptic focus is highly meaningful. In this talk, we would like to introduce a robust method for EEG source localization based on a new multiway temporal spatial-spectral (TSS) analysis via spectral graph theory and tensorial blind source separation. In particular, instead of using the temporal behavior of sources (i.e., time variable), we apply the graph wavelet transform (GWT), which is one of the most powerful vertex-frequency tools for graph signal processing (GSP), to the space variable in order to exploit information of the spatial domain. Numerical experiments on both simulated and real data are carried out to evaluate the model performance of the TSS analysis in comparison to space-timefrequency (STF), space-time-wave-vector (STWV) analysis.
Speaker: Mr. Le Trung Thanh, AVITECH
Time: 15:30, Tuesday, September 03, 2019
Venue: G2-315, 144 Xuan Thuy, Cau Giay, Hanoi
Le Trung Thanh received B.Sc. and M.Sc. degrees in Electronics and Telecommunications from the VNU University of Engineering and Technology (VNU-UET), a member of Vietnam National University, Hanoi (VNU) in 2016 and 2018 respectively. He is now pursuing his Ph.D. study at the University of Orleans, France. His research interests include signal processing, subspace tracking, tensor analysis, and system identification.