| Equivalent dipole localization is a brain electrical source localization method widely used in electroencephalogram inverse problem (EEG IP) research. EEG IP has unique advantages in brain function research, cognition research and brain disease detection. This paper discussed the sensitivity relationship between electrical conductivity distribution and equivalent dipole localization by simulation of EEG IP. On the basis of current construction methods of head models, four-layer sphere model was constructed in this paper according to asymmetry prosperities of brain tissues. Using linearity adding theory, spouse theory, mirror method and space rotation transform, the analysis solution comes from widely used three-layer sphere model. The result was verified by calculation with FEM software. Given reasonable initial value and restrict condition, equivalent dipole localization was worked out by using Sequential Quadratic Programming (SQP). Based on the simulation method above, two propositions were studied as following: 1. The respondence of equivalent dipole localization when the eccentric sphere with high electrical conductivity exists. 2. The influence of skull electrical conductivity on equivalent dipole localization. The result proves that electrical conductivity distribution of brain tissues has significent influence on equivalent dipole localization. Consequently it is the foundation of improving localization precision to obtain the correct electrical conductivity and geometry information of brain tissue. Here we built up an electrical impedance tomography (EIT) system, and the images of model testing were acquired successfully, which proves the feasibility of application. Furthemore,advices of next improvement of applications on real brain tissue are also discussed. On the basis of calculation and simulation above, the result shows distribution of brain electrical conductivity has significent influence on equivalent dipole localization. Advanced brain function research will be in the nature of being promoted greatly by integrated technique of EEG IP and EIT. |