Dear all,
I performed tract-based statistical analysis (TBSS). After the randomise, how can I extract the cluster and quantitative values? I can visualize the results with fslview, but I need quantitative values for correlation with clinical parameters.
Many thanks!
Celi
Hi Celi,
Cluster command is very useful. Use cluster -h for help.
Example to check for 95% significant results in the t-test 1 comparative. Results will be saved in the cluster_t1_95.output file:
cluster -i tbss_tfce_corrp_tstat1.nii.gz -t 0.95 --mm > cluster_t1_95.output
If you have more specific questions about it, do not hesitate to ask me.
Regards,
Alba
Hi Celi,
Cluster command is very useful. Use cluster -h for help.
Example to check for 95% significant results in the t-test 1 comparative. Results will be saved in the cluster_t1_95.output file:
cluster -i tbss_tfce_corrp_tstat1.nii.gz -t 0.95 --mm > cluster_t1_95.output
If you have more specific questions about it, do not hesitate to ask me.
Regards,
Alba
Dear Celi,
for the cluster identification, I also use "cluster" command under FSL.
For the atlas, I usually use the Anatomy toolbox for SPM, that allows me to obtain the clusters in an editable and easy to use format, like txt (I usually import the result file in Excel, and work on that).
In my opinion, it works better then the "atlasquery" command under FSL.
Best Regards,
Sirio
Hello everyone
i'm a TBSS beginner and i'm wondering How can i extract FA skeleton value after randomize step on TBSS
so basically i need to know the FA value of each subject or the mean FA of each group within significant areas highlighted on the skeleton
thank you beforehand
To Eya: fslmeants from FSL and/or mri_segstats from Freesurfer may help.
I am trying to do that and found this question very helpful. Here is my summary:
1. Use cluster to get an image of cluster indices and a table of p-values and peak as well as center of gravity coordinates.
2. Input the all_FA_skeletonised into fslmeants with the cluster index image as the --label option to get a mean FA per subject per cluster (it will be a matrix of values where rows are subjects and columns are clusters).
Thanks everyone!
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis