Multimodal electrical impedance tomography and electroencephalography imaging: Does higher skull conductivity resolution in EIT imaging improve accuracy of EEG source localization?

Abstract

ObjectiveUnknown conductivities of the head tissues, particularly the skull, is a major factor of uncertainty in electroencephalography (EEG) source imaging. Here, we develop a personalized skull conductivity framework aiming to improve the head models used in the EEG source imaging and to reduce localization errors. MethodsWe employ Electrical Impedance Tomography (EIT) and convex optimization to produce high resolution skull conductivity maps that are subsequently embedded in the EEG modeling. ResultsFirst, we demonstrate through simulations that locally varying conductivity values of the skull can be estimated from EIT measurements. Second, we show how the choice of the skull conductivity resolution of the EIT imaging affects the EEG source reconstructions. ConclusionsEIT estimated conductivities can signicantly improve the source reconstructions, particularly in cortical areas under bones that exhibit high conductivity variations. SignificanceThis work acts as a steppingstone in defining a protocol for the preparation of patient-specic head conductivity models that are essential for accurate examination and systematic monitoring of the brain activity via EEG.