Multimodal neuroimaging integrating functional magnetic resonance Imaging and electroencephalography.

Abstract

Functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) are two widely used neuroimaging modalities with complementary merits and limitations. FMRI has low temporal resolution but high spatial resolution, while EEG has low spatial resolution but high temporal resolution. The researches included in my dissertation aim at developing an fMRI-EEG integrated multimodal functional neuroimaging approach with significantly enhanced spatiotemporal resolution relative to fMRI or EEG alone. Towards this goal, i) we have established the capability to simultaneously record EEG and fMRI signals; ii) we have mathematically modeled the interactions between stimuli (or tasks), synaptic currents and BOLD responses, and have developed a model-driven linear regression method for quantifying BOLD fMRI signals to characterize event-related electrophysiological responses; iii) we have developed two novel algorithms based on an adaptive Wiener filter and the Twomey regularization, respectively, for the fMRI-constrained cortical current density imaging; iv) we have employed the developed fMRI-EEG technique to investigate the cortical visual pathway and the cortical substrates responsible for the bilateral visual integration. The outcome of these researches is encouraging. Significant progresses have been achieved in both technical and fundamental aspects of the fMRI-EEG integrated neuroimaging, which holds potential to open a unique window to non-invasively image and investigate electrophysiological brain activity, as well as its relationship to cerebral hemodynamics.