Anisotropic equivalent conductivity tensors for bioelectric modeling of partial volume effects in cerebrospinal fluid spaces

Accurate modeling of bioelectric propagation within the head is necessary for precise electromagnetic source localization. We present here a new approach for modeling spaces fractionally composed of grey matter and cerebrospinal fluid. Using information about the orientation of the cortical surface, we construct anisotropic conductivity tensors to model the partial volume effects frequently present within sulci. Our results indicate that the CSF region has significant effect upon the resulting model, and can more accurately be accounted for using anisotropic conductivities. We validate our method using a whole brain statistical error analysis of the signal topography and magnitude.