Pain FMRI studies are improved by slice-wise removal of cardiac noise
Vogt K, Ibinson J, Small R, Schmalbrock P
The Ohio State University, The Ohio State University
This study directly measured subjects pulse plethysmograph during functional MRI experiments with painful stimulation, during which cardiovascular changes are significant. This measured cardiac data was used to remove the cardiac noise in the time-ordered image data from each image slice by linear regression. Subsequent statistical analysis of activation coinciding with the pain stimulus paradigm showed 23% better fit of the general linear model, as measured by the average change in the adjusted coefficient of determination across all voxels. Areas exhibiting strong correlation to cardiac noise were those associated with large cerebral vessels, consistent with previously published maps.