[SOLVED] Fixelcfestats problem with fd data

[rgrazioplene]

Hi again, I am using fixelcfestats to run some regressions with diagnosis and age (respectively), looking at both fd and fc. The results for the fc look reasonable, but the fd results do not appear to be at all constrained by connectivity. For example, the uncorrected p-value maps for fc look like this:

whereas the uncorrected p-values for fd look like this:

The code I used for each was exactly the same, and the analyses appeared to run without difficulty

fixelcfestats -force -negative -nperms 2000 input_fd_146.txt analysis_fixel_mask.msf design_matrix_dx_sex_age_146.txt contrast_matrix_dx_ones.txt tracks_2_million_sift.tck dx_146_fd

fixelcfestats -force -negative -nperms 2000 input_fc_log_146.txt analysis_fixel_mask.msf design_matrix_dx_sex_age_146.txt contrast_matrix_dx_ones.txt tracks_2_million_sift.tck dx_146_fc_log

fixelcfestats produces quite a bit of output that I’ve had a hard time finding documentation on, so I’m at a loss as to how to diagnose this issue. Has anyone else encountered this kind of problem? Anything I should be checking or looking for in the case of this kind of output?

[rgrazioplene]

Though I feel a bit sheepish, I thought I’d go ahead and answer this myself for posterity. Due to an overwrite error, my fod2fixel analysis (Step 9) generated empty images. Down the line (fixelcorrespondence) this assigns the value of the analysis_fixel_mask to all subjects’ data–no wonder my model results looked like pure noise for fd!! I didn’t pick up on it immediately because the fd images were not technically empty themselves (they were just copies of the fixel analysis mask).