Network-based techniques in EEG data analysis and epileptic brain modeling

Abstract

We discuss a novel approach of modeling the behavior of the epileptic human brain, which utilizes network-based techniques in combination with statistical preprocessing of the electroencephalographic (EEG) data obtained from the electrodes located in different parts of the brain. In the constructed graphs, the vertices represent the "functional units" of the brain, where electrodes are located. Studying dynamical changes of the properties of these graphs provides valuable information about the patterns characterizing the behavior of the brain prior to, during, and after an epileptic seizure.