An Event-related Analysis for Identifying Networks Integrated by P300 Simultaneously Measured with EEG and fMRI

In this study, P300 that induced by visual stimuli was examined with simultaneous EEG/fMRI. For the purpose of combine the best temporary resolution with the best special resolution together to estimate the cortical network, event-related analysis contributed to this methodological trial. A 64 channel MRI-compatible EEG amplifier (BrainAmp: made of Brain Production GmbH, Germany) was used in the measurement simultaneously with fMRI scanning. The sampling rate of raw EEG was 5 kHz, and the MRI noise reduction was performed. EEG recording synchronized with MRI scan by original stimulus system, and an oddball paradigm (3 letters presentation: 'T'=Standard with 56% appearances proportion, 'C'=Target with 22% appearances proportion, 'P'=Novel with 22% appearances proportion) was performed. Presentation time of visual stimulus is 240ms and inter-stimulus interval (ISI>650ms) is variable. Every subject underwent a 1344s session, there are eight 156s stimul-blocks which including 26pairs of stimulation and eight 12s interval. MRI scan interval (3s) is synchronized to inter-stimulus interval (3s) in the whole session. In order that activations of P300 can be scanned by each slice scanning-phase, the appearance of visual stimuli shifted from start to the end of each MRI scan randomly. After P300 segmentation, the timing of P300 was exported to event-related analysis of fMRI data with SPM99 software. In the 7 subjects group study, the significant activations appear in the inferior occipital gyrus of both cerebrum, superior and medial frontal gyrus of both cerebrum, precentral gyrus of both cerebrum, left superior temporal gyrus, and left inferior parietal lobule. Our method successfully demonstrated a cortical network that related with P300 event during visual cognition. The spatial information of the cortical regions provides a possible function of P300 that activated at the frontal and parietal sites as an attention control for Logical Judgment process in the visual cognition. By using a simultaneous fMRI and EEG, even a small phase synchronization (P300 event-related) is good enough to extract the functional relevant networking of regions, not only in the superficial, but also deeper brain regions. This method could be feasible for identifying networks that are integrated with event.