Spontaneous EEG Rhythms and their fMRI Correlates

A major limitation of the routinely used clinical EEG lies in the ambiguous assignment of measured brain potentials to anatomic structures. By means of simultaneous EEG-fMRI the non-invasive imaging of structural correlates of EEG rhythms has become feasible, thus improving the interpretability of EEG findings. During MRI acquisition, however, large artifacts are induced in the EEG signal, making the physiological EEG unrecognizable. We employed a mathematical algorithm for artifact removal and validated this algorithm by means of visual evoked potentials (VEPs) recorded during and between MRI acquisition periods. We also showed the feasibility of MR imaging of physiological background rhythms like Alpha rhythm, functionally connected with the visual system, and sensorimotor Mu-rhythm. The MRI blood oxygenation level-dependent (BOLD) signal was negatively correlated with posterior Alpha rhythm in occipital areas, indicating deactivation in a metabolic sense. This result was confirmed by near infrared spectroscopy (NIRS), an optical method for monitoring local changes in deoxygenated hemoglobin concentration. Positive correlations between BOLD signal and Alpha activity in the thalamus indicate its generator role for this rhythm. For the 10- and 20-Hz components of the central Mu-rhythm we identified different cortical and subcortical representations. MR correlates of pathological EEG rhythms were investigated in patients with hepatic encephalography. Inevitable changes of vigilance during EEG-fMRI measurements, affecting EEG as well as BOLD signal, make simultaneous monitoring of vigilance necessary. A methodological approach is in progress with first results available.