Based On A Realistic Head Model Of The Eeg Forward And Inverse Methods Of Research

Electroencephalogram (EEG) is a measurement of the time-varying potential differences between electrodes on the scalp. It is used in many applications where the spatial and temporal aspects of electrical brain activities are needed. In the study of EEG , solving the inverse problem will provide us with information about the location, intensity and distribution of the neural current sources. Yet, the base of the inverse problem is the forward problem. The forward problem includes the activity source model, the head model and the mathematical model.In this thesis, the forward problem based on the boundary element method (BEM) and the inverse problem on the dipole fitting and equivalent source layers mapping has been studied. The main contents are as follows:1. The study of the EEG forward problem: using the BEM to solve the forward problem in the source model of the dipole and charge. In the three-shell model, we compared the scalp potential estimated by BEM to that analytically calculated, and the comparison result validated the precision of the BEM. We found that the precision of our BEM was similar to those in the recently literature. Finally, we realized the modified BEM in the literature. It may save great time in the debugging phase of the inverse problem.2. The study of EEG inverse problem: using the evolution strategy (ES) arithmetic for the location of the instantaneous dipoles. On one side, for the ES only needs the mutation to generate the offspring and keep the best offspring, so it can enhance the computational efficiency. On the other hand, we used the combination strategy in the multiplicities dipoles to increase the astringency of the arithmetic and the precision of the location. In the equivalent source layers mapping (ESM), we firstly achieved the equivalent charge layers mapping (ECM) in real head model, compared with the equivalent dipole layers mapping (EDM), the imaging resolution is better.