I don't understand. How did you resample your data? If your image could overlaid on 2mm brain, it means they have the same size. You could follow this link https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=fsl;4baa23b3.0709 to up-sample your image, which will definitely be able to be overlaid on 1mm brain.Good luck : )On Sat, Apr 8, 2017 at 10:42 PM, DTI students from China <[log in to unmask]> wrote:Thanks for your help! I had solved the problems causing to a subtle mistake in my processing.
And may I consult you about another problem about FSLview?
I found that the atlas viewed in FSLview was labeled as the "aligned anatomical" coordinate space and when could be overlaid by the MNI_152_T1_brain_2mm image. Then the 2*2*2 atlas was resampled into 1*1*1, the labeled coordinate space was still the "aligned anatomical", however the image could not be overlaid by MNI_152_T1_brain_1mm image any more.
I am looking forward to your reply if your time was permitted.
At 2017-04-02 10:57:18, "DTI students from China" <[log in to unmask]> wrote:
Thanks a lot for your precise description of my problems I met. And I have another questions that seem like foolish~~
If I want to find the correlation between grades of scales and mean FA of individual mean FA of significant voxels, could I directly extract the mean FA of significant voxels at the skeleton?
fslmaths tfce_corrp_stats -thr 0.95 -bin thr_mask
fslmeants -i all_skeletonised -m thr_mask -o individual_mean_FA.txt
Or, it is better to deproject the voxels in mean_FA_skeleton to individual diffusion space.
tbss_deproject thr_mask 2 -n
get the mean_FA of individual separately?
But I am not sure that extracting the FA of skeleton directly is right.
For the second question, the atlas I uesd to label clusters is 2*2*2 in MNI space and was resampled into 1*1*1. However when I used atlasquery and autoaq tool within FSL, it can not work. Is there any solutions?
Thanks again for your patience.
At 2017-04-02 01:41:16, "Guangyu Zhou" <[log in to unmask]> wrote:
I guess you are talking about labelling significant clusters tfce_corrp_stat1 and tfce_corrp_stat2, not different ways of generating those stats images. As long as the atlas in the software you were referring is also in MNI52 standard space, this should be fine. The total number of significant voxels should be the same, no matter which atlas was used for labelling. The number of significant voxels within each brain region could be slightly different which depends on which atlas you use.On Sat, Apr 1, 2017 at 12:32 AM, SUBSCRIBE FSL Shufei Zhang <[log in to unmask]> wrote:Dear All,
I am new with FSL and met a problem when reporting the results.
Because I want to report the results of TBSS through the atlas different from the bulit-in atlas of FSL. I used another software to report it. My problems are described below.
Is it permittable? when I used another software to locate the results in the atlas and the number of two-significant-cluster voxels were aboult 5000 but only the 227 voxels significant at the tfce_corrp_stat2 were found by fslstats. I know tbss_skeleton was used to represent local maximal and it is right to report the results directly by cluster terminal?
Thanks for your petience and if I can not describe clearly, contact me plz. I really appreciate FSL software that helps me a lot in DTI data analysis.
Yours:Shufei Zhang