Gradient Artifact Removal in Co-registration EEG/fMRI

In this work we present a novel method to remove the gradient artifact in co-registered EEG/fMRI, of wide applicability and able to preserve EEG integrity. It is based on an iterative subtraction, but it improves upon exiting methods since for each slice the amplitude of the gradient templates to be subtracted is optimally adjusted by linear regression. This avoids the need of adaptive noise cancellation, or principal component analysis. Performance of this algorithm was first evaluated on both in silico and real data. Results indicate that few iterations are needed for a successful gradient removal, without compromising frequency content of EEG.