 |
 |
Hi Hitoshi This question is correlated with another question "Freesurfer (surface to surface) tracking with probtrackx2". I did probtrackx2 using surface (seed=lh.LargeSTG.gii, exclusion=lh.pial.gii, target=lh.Broca.gii, reference=nodif_brain_mask_diffusion.nii.gz) but result was not good. I got no fdt_paths.nii.gz file. Do you mean that there was no output or that the output fdt_paths was blank (all zeros)? If there was no output then you may want to check that you’ve used the —opd option in probtrackx2. If the output is there but it’s blank, then either something’s wrong with the transformations to and from diffusion space, or all streamlines were genuinely rejected, in which case try to relax the exclusion/inclusion criteria and look at the results to try and figure out what might have happened. So I used "surf2volume surface.gii orig.nii.gz surface.nii.gz freesurfer" and got surface.nii.gz files. When I checked all files in fslview, everything was correct. Did you overlay it with a brain that is in the same space as the seed volume used in probtrackx2? Make sure you do that and if it works then you can be sure that the surface file is correct. In addition I did probtrackx2 using surface.nii.gz (reference=nodif_brain_mask.nii.gz) and result was maybe nice. Question1 I don't understand why probtrackx2 ( using surface.gii) did't work correctly. I thought that the nodif_brain_mask_diffusion.nii.gz file was not correct. I made the nodif_brain_mask_diffusion.nii.gz file (using flirt program, reference=str2diff.mat). I don’t think it’s a problem with the mask. It might be that sampling from the surface decreases your sensitivity too much in this particular case. Did you use exactly the same exclusion surface here or did you change that to be a NIFTI volume too? Question2 I have another question. If I want to change surface.gii to surface.diff.nii.gz, how do I use? Can I use same flirt program, reference=freesurfer2fa.mat, using a nodif_brain.nii.gz? No clear what you’re asking - seems to me that your surface was projected correctly into NIFTI volume in seed space, so that’s unlikely to be an issue… Saad Please give me a nice advise. Hitoshi 2016-03-17 19:35 GMT+09:00 Saad Jbabdi <[log in to unmask]<mailto:[log in to unmask]>>: Hi Hitoshi I am afraid I don’t understand what the problem is. Could you try to rephrase please? Thanks Saad > On 17 Mar 2016, at 05:39, Hitoshi Sasaki <[log in to unmask]<mailto:[log in to unmask]>> wrote: > > Dear Saad > > When I used .gii or .gii.nii.gz, in protbrackx2, results were different. > Tracking ( using .gii.nii.gz file) only got fdt_paths.nii.gz. > I wonder if nodif_brain_mask.nii.gz( in diffusion space) was not good, I don't know whether the way below was correct or wrong. > > To make a nodif_brain_mask_diffusion.nii.gz (-m,--mask Bet binary mask file in diffusion space) I used the following script. > > flirt -in $dti_subjects_dir/$D/bedpostx.bedpostX/nodif_brain_mask.nii.gz -ref $dti_subjects_dir/$D/bedpostx.bedpostX/nodif_brain.nii.gz -out $dti_subjects_dir/$D/bedpostx.bedpostX/nodif_brain_mask_diffusion.nii.gz -omat $dti_subjects_dir/$D/bedpostx.bedpostX/xfms/str2diff.mat -bins 256 -cost corratio -searchrx -90 90 -searchry -90 90 -searchrz -90 90 -dof 12 -interp trilinear > > Is it correct?? struct2fa.mat is correct? > > and > > If I want to change surface from .gii to diffusion, Can I use flirt, freesurfer2fa.mat, nodif_brain??