Surface ROI question

AFNI Message Board
1

So I know this might be a long-shot with Ziad gone but I’m trying to pull out just one cluster out of two that I found when comparing two groups using 3dttest++. I used 3dcalc to pull out values above a certain threshold and want to just pull out the two resulting clusters individually to use them as a mask for 3dmaskave. I’ve tried drawing an ROI around them and using 3dcalc/3dresample to no avail. I did also try SurfClust but I’m not certain how to use it exactly, I’ll post the command I’m using for that below.

SurfClust -spec ./fsaverage/SUMA/fsaverage_lh.spec -surf_A ./fsaverage/SUMA/lh.pial.asc -input lh.Groups3_2.gii 0 -rmm 2 -prefix clust -thresh_col 1 -thresh 3.00 -out_clusterdset

I believe the number after the input filename means the sub-brick to use, thresh_col means the sub-brick for the thresholding (sub-brick for t-stat probably), thresh means the threshold value (so t-stat threshold probably). Using various calls of this I haven’t been able to reproduce the results of 3dttest++. As you can see, I’m working with freesurfer data originally which is what the 3dttest++ results agree with, so I would imagine SurfClust would produce similar output.

Even if I did get SurfClust to work, I’m not sure how the results would translate to something 3dmaskave could use. Any help is very appreciated, I’m sort of running in circles with this.

Thank you,
Taylor H.

2

For both cases, drawing a surface ROI or “SurfClust’s -out_roidset”, I believe you should convert these abbreviated lists of nodes to full lists with something like this:

the prefix option has to be first here

ROI2dataset -prefix testout2 -input testroi.niml.roi

Alternatively, I think SurfClust -out_fulllist will work too.

3

I did find that command, and I’m able to output ROIs from SurfClust but I’m trying to pull out clusters manually from a file I’ll link here where there are two areas that were found to be very significantly different between groups. I’d like to separate the two clusters from this file if possible.

https://drive.google.com/file/d/0BwsiRctbBgfBWTNkOVBJbHVteFE/view?usp=sharing