Beyond localization: Strategies to analyze simultaneously acquired EEG and fMRI data in pain research

Simultaneous measurement of fMRI and EEG promises new insights into the topography and temporal dynamics of cortical processes and a more precise understanding of the relationship between the two measures. Together with a detailed behavioral analysis, this allows for a better elucidation of cortical pain processing. We applied painful stimulation to 6 healthy subjects and acquired fMRI simultaneously to an EEG measurement. After stimulation, all subjects rated different stimulus properties and the individual situation in the scanner. By applying a principal component analysis to the behavioral data, 2 main factors were extracted, the 1st (F1) reflecting the physical properties of the stimulus, the 2nd (F2) the subject's emotional situation in the scanner. Stimulus correlated BOLD effects were found in the primary and secondary somatosensory areas (SI and SII), the operculum, the insula, the supplementary motor area (SMA), the SMA proper, the cerebellum, and posterior parts of the anterior cingulate gyrus (ACC). In all subjects, neuroelectric source imaging revealed a contralateral source in the ACC and in all but one subject in SI. Sources in SII could be fitted in 3 subjects. While the second order fMRI analysis using F1 as predictor pointed to a cortical network, coarsely comparable to the first order analysis, using F2, the network was restricted to limbic and motor areas. Using dipole strength of the modeled source in SI as predictor, the network was restricted to SI. Over the course of the experiment the evoked BOLD signal change decreased in SI. Within the stimulation block, it decreased in SII from the first to the second third followed by an increase in the last third. In contrast, dipole strength of the ACC source decreased linearly over the experiment. Conclusion: The simultaneous measurement of pain evoked activity by EEG and fMRI results in concordant localizations in sensory and limbic structures. Differential habituation effects for sensory and limbic areas can be shown by fMRI and dipole analysis. The integration of the data reveals a close relationship of the two measures in primary sensory areas. Supported by the Deutsche Forschungsgemeinschaft (Fl 156/25-1)