December 14, 2010 16:0 WSPC/S0219-6352 179-JIN 00251 Journal of Integrative Neuroscience, Vol. 9, No. 4 (2010) 453–476 c  Imperial College Press DOI: 10.1142/S0219635210002512 EEG-fMRI INTEGRATION: A CRITICAL REVIEW OF BIOPHYSICAL MODELING AND DATA ANALYSIS APPROACHES M. J. ROSA ∗ , J. DAUNIZEAU ∗,†,‡ and K. J. FRISTON ∗ ∗ Wellcome Trust Centre for Neuroimaging, Institute of Neurology, University College London, United Kingdom † Laboratory for Social and Neural Systems Research, Institute for Empirical Research in Economics, University of Zurich, Switzerland ‡ j.daunizeau@fil.ion.ucl.ac.uk Received 1 August 2010 Accepted 17 September 2010 The diverse nature of cerebral activity, as measured using neuroimaging techniques, has been recognised long ago. It seems obvious that using single modality recordings can be limited when it comes to capturing its complex nature. Thus, it has been argued that mov- ing to a multimodal approach will allow neuroscientists to better understand the dynamics and structure of this activity. This means that integrating information from different tech- niques, such as electroencephalography (EEG) and the blood oxygenated level dependent (BOLD) signal recorded with functional magnetic resonance imaging (fMRI), represents an important methodological challenge. In this work, we review the work that has been done thus far to derive EEG/fMRI integration approaches. This leads us to inspect the conditions under which such an integration approach could work or fail, and to disclose the types of scientific questions one could (and could not) hope to answer with it. Keywords : Neuroimaging; information fusion; functional segregation; functional integra- tion; event-related; neurophysiology; data-driven; model-based; Bayesian analysis; model comparison. 1. Introduction The realization of any cognitive, motor or sensory process rests on cerebral dynamics and creates order in the bioelectric and hemodynamic signals measured with EEG and fMRI, respectively. To detect and interpret the relevant features of these signals, one typically describes processes at their own temporal and spatial scales. The main sources of scalp EEG signals are postsynaptic cortical currents associated with large pyramidal neurons, which are oriented perpendicular to the cortical surface [115]. EEG (and magnetoencephalography, MEG) is well suited to studying the temporal dynamics of neuronal activity, since it provides direct measurement of this activ- ity with millisecond precision. However, the scalp topology of measured electrical ‡ Corresponding author. 453