processing-scripts > Modifying the scripts for NYU_cocaine
Showing 1-3 of 3 posts
Results per page:
Nov 14, 2011  01:11 PM | Michal Pikusa
Modifying the scripts for NYU_cocaine

I have a question regarding the use of processing scripts on the NYU Cocaine dataset. I know that data from this set have been already preprocessed a little bit, so could anyone help me with editing the scripts so that they process the rest that haven't been done already? I tried modifying the scripts and running them, but to no avail. I know that since there is no anatomical scan, the anatomic processing should be deleted. Also, should i leave registration? I know that the volumes are already registered to MNI 3mm, but does any additional processing from the scripts affect it? I don't really know where to put the mask files and the motion correction file, and how to use them properly, since the structure of the directories is a little bit different due to absence of anatomical scan (no segmentation)

I have worked with the other NYU dataset and got it to work perfectly, so the scripts are working, but I can't get them to work with this particular dataset. If anyone could point me to what I should modify in the scripts to run them on NYU Cocaine, I would really appreciate it.
Nov 15, 2011  09:11 AM | Michal Pikusa
RE: Modifying the scripts for NYU_cocaine
I managed to run the scripts properly and tested a couple of seeds with RSFC, and the resuts look ok, but I want to make sure I modified the scripts in the right way, so that my results are not false.

I removed every process from the functional preprocessing script up to motion correction (since slice-timing correction, removal of first timepoints, and reorientation have been done), then I left motion correction and all that followed (excluding registration, since it's been done), and it went well. I omitted anatomical preprocessing and segmentaion, since there are no anatomic volumes. Then, I created a segment directory and copied the mask files that were included in the dataset to that directory and then I ran nuisance signal regression script, and it went well, too. Afterwards, I ran RSFC script without any modifications and I got my correlation files, which is good.

Is this the right approach?
Nov 21, 2011  09:11 AM | Maarten Mennes
RE: Modifying the scripts for NYU_cocaine
Hi Michal,

Sorry for the slow response.

Yes, the approach you took sound right. Very good that you were able to work it out and thanks for posting it back here. This information will be very useful for others too.