Simulated electroencephalography (EEG) source localization using integrated meromorphic approximation

Abstract

Epilepsy is a chronic brain dysfunction in which neurons and neuronal network malfunction cause symptoms of a seizure. A seizure is an abnormal electrical discharge from the brain appearing at a small area of the brain. The seizure affected zone loses its normal task abilities and might react uncontrollably. Electroencephalography (EEG) is one of the useful instruments in diagnosing many brain disorders like epilepsy. This non-invasive modality is used to localize brain regions involved during the generation of epileptic discharges. At present, many quantitative methods for identifying and localizing the epileptogenic focus from EEG have been invented by scientists around the world. Under quasi-static assumptions, Maxwell’s equations governing the spatial behaviour of the electromagnetic fields lead to Partial Differential Equations (PDE) of elliptic type in domains of R3. This thesis presents a new method based on integrated new EEG source detection, Cortical Brain Scanning (CBS) with meromorphic approximation to identify the sources on the brain scalp, which have highly abnormal activities when a patient is having a seizure attack. Boundary measurements for meromorphic approximation method are considered as isotropic and homogeneous in each layer (brain, skull, and scalp). The proposed method is applied on simulated and published EEG data obtained from epileptic patients. The method can enhance the localizations of sources in comparison to other methods, such as Low Resolution Brain Electromagnetic Tomography (LORETA), Minimum Norm Estimation (MNE), and Weight Minimum Norm Estimate (WMNE), coupled with meromorphic approximation. Standard validation metrics including Root Sum Square (RSS), Mean Square Error (MSE), and Receiver Operating Characteristic Curve (ROC) are used to verify the result. The proposed method produces promising results in enhancing the source of localization accuracy of epileptic foci.