Reduction of ballistocardiogram artifact in the simultaneous acquisition of auditory event-related potentials and functional magnetic resonance images

The combination of electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) is now being used to analyze brain function. However, the ballistocardiogram artifact (BA), which is caused by the pulsatile motion related to the heart beat, dominates the EEG signal in the MR environment. The objective of this research is to reduce these MR-induced artifacts in order to extract useful information from the EEG recordings, specifically auditory event-related potentials (ERPs). The BA was reduced through the use of an algorithm which creates an adaptive BA template by median-filtering the previous 10 artifacts in the EEG signal. The continuously updated template is then subtracted from each BA in the EEG signal. Results are reported for five subjects, and it was found that the results for the mismatch negativities (MMNs) were comparable to those that have previously been reported for studies using auditory frequency deviants. The methods used to reduce the BA in this study demonstrate that the acquisition of auditory ERPs along with the subsequent MMN component can validly be obtained in the MR-environment.