Hello everyone,
I'm new to FreeSurfer and this question might sound a bit stupid: I'm trying to get the volumetric measures, surface area and cortical thickness of brain regions of three groups of subjects (patient group 1, group 2 and healthy controls), and I was wondering whether the morphometric variance across subjects (gender, shape of brain, etc) unrelated to the neurological condition has been corrected automatically within the recon-all pipeline using the default options.
That is, given that the groups are age-, gender- and otherwise matched, after segmentation and calculation of these values with FreeSurfer, the morphometric differences we are looking at could be attributed to the neurological condition itself, instead of other factors.
From what I read, the step Talaraich which computes the affine transform from the orig volume to the MNI305 atlas (https://surfer.nmr.mgh.harvard.edu/fswiki/recon-all#Talairach.28-.3Cno.3Etalairach.29) seems to be one of the steps that does what I need, and we are therefore safe using the default recon-all for pre-processing and extraction of morphometric data. Otherwise, please let me know what I could do.
Thanks very much for your information and help in advance!
Article Impaired perfusion and capillary dysfunction in prodromal Al...
Article Capillary dysfunction is associated with symptom severity an...
hope you will find inspiration and information in those papers
Hi Jie Mei,
"... I was wondering whether the morphometric variance across subjects (gender, shape of brain, etc) unrelated to the neurological condition has been corrected automatically within the recon-all pipeline using the default options."
The answer to that question is no. The default recon-all pipeline builds surface meshes that resembles brain tissue as close as possible. This means that subject specific variance is not removed, but estimated.
" That is, given that the groups are age-, gender- and otherwise matched, after segmentation and calculation of these values with FreeSurfer, the morphometric differences we are looking at could be attributed to the neurological condition itself, instead of other factors."
If you average morphometric data over groups that are e.g. age-matched. Then it is likely that you average out effects of age. Another option would be to pass e.g. age as a covariance factor in your statistical tests. That would also allow you to assess the strength of these factors.
Hope this helps.
Wouter
Hi,
I would strongly recommend doing a quality check for the output data. If you observe anything beside brain tissue, edit the pial surface, and check for the boundary.
Then, regards with the head variation, you could normalise with either TBV or intracranial volume. If you want to provide covariates in the group analysis, either you do in the software (FS) or you could extract all the morphometric measures and do a GLM with the independent variable as covariates eg. sex, age.
Or else, first check the groups, how it differs for intracranial or TBV, based on that you could add as covariates (if the volume significantly differs between groups).
And, even though the groups are well-matched, its more likely they differ with head size.
I hope this helps.
- Badges
- Science topic
- Similar topics
- Biological Science
- Neuroscience