A finite difference model of electric field propagation in the human head: Implementation and validation

Abstract

It is shown that a realistic finite-difference computational model for the propagation of electrical fields within the human head can be effectively implemented on a massively parallel computer. Potentially, the model will have not only anatomical accuracy, but also will be applicable to individual subjects automatically using their MRIs and/or CTs. This model has two further advantages. First, it yields the electric field distribution for all points within the head, not only at the scalp. Second, it allows the calculation of electric fields resulting from complex generator configurations without additional processing time, simply by placing additional sources and sinks at locations determined by the individual's MRI. Analytical methods can also model generators as dipole layers with complex geometries as the sum of many individual dipoles. However, the calculation increases proportionally with the number of modelled dipoles, and the placement of dipoles is problematic because the brain is not spherical. The model requires further development before it can be usefully applied to understanding the generation of human EEG.<>