This describes how to process relative cerebral blood flow (relCBF) data produced by the Siemens scanner using their pulsed ASL (PASL) protocol in the FreeSurfer environment, including registration and group analysis. This is sequence is usually called "ep2d_pasl". If you are not using this sequence, this information is NOT for you. This sequence will produce a several series, one of which will have "relCBF" as the series description. It will have one time point.
There are two programs that will be used here: rcbf-prep and rcbf-merge, each has its own help which you can see by running with --help.
The output relCBF will be in mL/100g/min.
The analysis will be done in two or three stages. First, analyzing each individual subject with rcbf-prep. Second, merging all the individual results with rcbf-merge. Finally, if you want to perform group analysis, then you will use mri_glmfit.
Individual Analysis (rcbf-prep)
For each subject, run the following command:
rcbf-prep --s subject --rcbf relCBF.nii --o subject-rcbf
* "subject" is the FreeSurfer subject.
* relCBF.nii is the Siemens output file without any scaling. This can be any format, even dicom.
* subject-rcbf is the output directory
This will create output directory subject-rcbf. It will modify the input volume to convert to mL/100g/min by subtracting 2048 and dividing by 5. It will register the input relCBF with the FreeSurfer anatomical using bbregister and resample to the surface (individual and fsaverage) and MNI305 as well as create an roi.dat file with relCBF averages over the given set of ROIs.By default, the ROI set is from $FREESURFER_HOME/FreeSurferColorLUT.txt, which has a huge number of ROIs. If you want to use a different set, copy and modify FreeSurferColorLUT.txt, and pass your new ROI table with --roitab. Note that aparc+aseg.mgz is used by default. This command should take about 5 min to run.
When complete, check the registration with:
tkregister2 --mov subject-rcbf/rcbf.nii --reg subject-rcbf/register.dat --surf
Merging Individual Analyses