Hi experts,
I am trying to correct for partial volume in ASL data that has been
processed using Ze Wang's ASL Toolbox. However, I would like to use
the partial volume correction (--pvcorr) option from oxford_asl.
I was wondering which step of preprocessing it is recommended to do
partial volume correction? My understanding from Chappell et al. 2011
paper on PV corr is that it is done simultaneously with the
tag-control subtraction? Therefore, should I be doing this on the
perfusion weighted images or on the raw unsubtracted timeseries?
Also, I am regressing out WM and CSF signals (from thresholded FAST
segmented masks) from the unsubtracted ASL timeseries, so perhaps
partial volume correction should be done before that process?
Furthermore, how does spatial smoothing affect partial volume correction?
Please advise.
Thanks so much,
Ishtiaq
|