Dear all,
I have some questions I could not find an answer to myself, I'd really appreciate it if anyone could help out.
1) How should recon-all be used to process all subjects' data (i.e. obtain the surface for each subject) before taking it to QDEC (or command prompt) for a group thickness analysis? The QDEC tutorial mentions a separate tutorial for the recon-all preprocessing part, but it is not at all clear which one that is. Is there more to it than just typing "recon-all --i file.nii –i file.dcm –subject bert –all" for each subject (assuming NIFTI structurals)?
2) Is mri_glmfit FreeSurfer's equivalent of the randomise command in FSL, i.e. it does nonparametric permutation-based inference based on the given design matrix and produces statistical maps? Or are there differences between the two in how the GLM is applied?
3) Freeview and tksurfer seem to overlap in scope - when would you use one rather than the other? Does it have to do with viewing structurals vs statistically-generated maps?
4) I see that points on reconall-generated surfaces are always referred to in tutorials as vertices rather than voxels. Since both represent the "unit of space", what exactly is the difference between them?
Many thanks in advance! Tudor
On 12/18/2012 06:48 AM, Tudor Popescu wrote:
Dear all,
I have some questions I could not find an answer to myself, I'd really appreciate it if anyone could help out.
- How should recon-all be used to process all subjects' data (i.e.
obtain the surface for each subject) before taking it to QDEC (or command prompt) for a group thickness analysis? The QDEC tutorial mentions a separate tutorial for the recon-all preprocessing part, but it is not at all clear which one that is. Is there more to it than just typing "recon-all --i file.nii –i file.dcm –subject bert –all" for each subject (assuming NIFTI structurals)?
To have it run through to the end, that is all you need (note that if file.nii and file.dcm represent the same volume, then only pass one). You will need to visually inspect and possibly manually fix your data. That is what the trouble shooting tutorial is about.
- Is mri_glmfit FreeSurfer's equivalent of the randomise command in
FSL, i.e. it does nonparametric permutation-based inference based on the given design matrix and produces statistical maps? Or are there differences between the two in how the GLM is applied?
mri_glmfit just performs the GLM analysis, ie, given an input y and a design matrix X and contrast matrices C1, C2, etc, it will fit the model to the data, compue beta's and residual variances, and test the contrasts, and computes parametric statistics. mri_glmfit-sim computes stats based on permutation or simulation stats.
- Freeview and tksurfer seem to overlap in scope - when would you use
one rather than the other? Does it have to do with viewing structurals vs statistically-generated maps?
They do overlap a lot and there is on one case where you would use one over the other. freeview is newer, and tksurfer will eventually be phased out.
- I see that points on reconall-generated surfaces are always
referred to in tutorials as vertices rather than voxels. Since both represent the "unit of space", what exactly is the difference between them?
Conceptually, not much. "Vertex" is a term commonly used in mesh models in which the distance between the vertices is not fixed. "Voxels" comes from imaging. Really, they just represent different communities.
doug
Many thanks in advance! Tudor
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas N Greve -
Tudor Popescu