Dear FSL experts,

 

I have a question about a pcASL case that went wrong. During a pcASL scan of one subject, the scanner stopped three times, and I restarted the scan immediately. Each scan has an either odd or even number of volumes, I assume that for a complete control-tag scan one must have an even number in order to be able to merge the different scans and keep the control-tag sequence. Since the scanner stopped a few times I used fslsplit -t to split each of the 3 scans in order to delete the last volume or the last pair in case they are not completed. I deleted either one or two volumes from the end of each scan to get to an even number in order to merge the files. After splitting the original data, I then used fslmerge -t vol* to put each of the 3 sets together individually and then again fslmerge -t to put them all 3 data sets together in one set of data. When I run the FEAT analysis it looks like the control and tag sequence was swapped and I  seem to see a lot of random activation (compared to the analysis I did for each of the individual scans) or no activation. I have a few questions that I don’t seem to find answers to:

 

1. Is it ok to put pcASL files together from the same subject in this way?
2. Does fslsplit -t split volumes from 0 to N in time, or it does it in a different way?
3. Is fslmerge -t vol* adding the volumes alphabetically after their numbers ie vol0000, vol 0001, vol 0002…vol00nn or it does it in a different order?
4. Are there any other ways of doing the operation I am trying to perform?

 

I look forward to hearing from you!

 

Kind regards

Eduard Vasile