Separation of Nonstationary EEG Epileptic Seizures Using Time-Frequency-Based Blind Signal Processing Techniques

Abstract

Epilepsy is a neural disorder in which the electrical discharge in the brain is abnormal, synchronized and excessive. Scalp Electroencephalogram (EEG) is often used in the diagnosis and treatment of epilepsy by examining the epileptic seizures and epileptic spikes. By modeling the signal acquired at each electrode of the EEG measurement system as a linear combination of source signals generated in the brain, we can apply Blind Source Separation (BSS) techniques to separate the seizures from other signals. Alternating Columns - Diagonal Centers (AC-DC) and Second-Order-Blind Identification (SOBI) are well-known BSS algorithms and have been previously applied to the separation of seizures. However, the seizure signals in new-born babies exhibit nonstationary second order statistics. In this paper, we concentrate on applying two time-frequency (TF) based algorithms: TF-SOBI and TF-UBSS to seizure separation. These algorithms are more appropriate for analyzing nonstationary signals and have not been previously applied to studies of EEG-based seizures.