Since neural BOLD responses were found in the white matter (WM), we hypothesized that using CSF and non-brain areas (NBA) as noise source areas, rather than WM and CSF, to denoise fMRI data with aCompCor, will positively affect the results of task fMRI studies. This hypothesis was tested by comparing the results of processing 2 task fMRI datasets once denoised with WM and CSF and once with CSF and NBA as noise source areas. Our results showed that denoising with aCompCor using CSF and NBA leads to higher effect sizes and a better overlap between individual activation maps.
This abstract and the presentation materials are available to members only; a login is required.