Hello,
I am having trouble getting significant results from a linear regression involving the mean FAs for every subject of an ROI deemed significant by TBSS and a continuous outcome score. I utilized the TBSS procedure to analyze the DTI data from neonates then randomised with age and sex covariates. Here is the code I used to extract a list of mean FAs for each subject for a ROI which I manually created based on the tracts TBSS called significant:
fslmaths all_FA_skeletonised.nii.gz -mul <ROI_mask> output_1
fslmaths output_1 -mas mean_FA_skeleton_mask.nii.gz -bin output_2
fslmeants -i all_FA_skeletonised.nii.gz -m output_2 -o <meanFA_list>
Then, I take the mean FA list and corresponding continuous outcome scores and use them to generate a linear regression in R. However, I get very poor results using this approach, and I'm not sure why.
If you have any suggestions for why this is not working, please let me know, but I also have a couple specific questions myself:
- Would making the ROI from the results of tbss_fill include non-significant regions? Should I take it instead from the direct result of randomise?
- In the linear regression in R, should I be adjusting for age and sex covariates again even though they were part of the randomise calculation?
- Does the fact that TBSS deem a region significant automatically mean that region is significantly correlated with outcome?
Thank you for your assistance,
Milan Parikh
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|