ABSTRACT
Recently, brain imaging techniques have been developed by numerous efforts directed at multimodal data fusion, which seeks to combine hightemporal resolution information, as can be provided by electromagneticbased techniques (M-EEG), with high-spatial resolution, as has been traditionally achieved by the use of hemodynamic-based neuroimaging methods such as positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) [1]. In particular, the EEG/fMRI fusion has been well known for more than a decade owing to the higher levels of information on brain activity that can be obtained by the simultaneous
CONTENTS
9.1 Introduction ................................................................................................ 195 9.2 Methods for M-EEG/fMRI Fusion ........................................................... 198
9.2.1 Correlation between BOLD and M-EEG Signals in Particular Bands ............................................................................. 199
9.2.2 Extract Common Spatial Signatures for M-EEG and fMRI .....200 9.2.3 M-EEG-Informed and fMRI-Informed Analyses ...................... 201 9.2.4 DCM-Based Generative Models .................................................. 203 9.2.5 Granger-Causality-Based Models ............................................... 204 9.2.6 Neural-Mass-Based Generative Models ..................................... 205
9.3 Significance of Fusion in Various Modalities ........................................ 206 9.4 Practical and Realistic M-EEG/fMRI Data Integration ........................ 208 9.5 Conclusions ................................................................................................. 212 References ............................................................................................................. 213
combination of these complementary modalities, and the significantly increasing capacities for carrying out these studies [2,3].
