Hi,
We are planning to use some of our own methodology (mainly segmentation routines) in the Freesurfer pipeline to make cortical thickness analysis between groups. Related to this, there is two things that would need clarification.
1. In case we would use our own surfaces for analysis, what surface properties would be needed by the registration software (mris_register) of Freesurfer, and, more precisely, exactly what curvature files? In the www-site of Graham Wideman, he presents a data flow chart that tells the input and output to the number of programs used in the analysis. He has concluded that the registration would need only the spherical surface file (rh.sphere or lh.sphere), but that can't be correct, right? Some curvature information needs to be used, but what? In the article 'High-Resolution Intersubject Averaging and a Coordinate System for the Cortical Surface' (HBM 1999) it is said that convexity of the surface can be used. But what is the property (and curvature file) used in Freesurfer?
2. Then we would of course need our own reference surface for registration, since we are analysing surfaces constructed differently from the default Freesurfer surfaces. Is there documentation on what data the reference surface (e.g. the rh.average.tif or lh.average.tif) contains? Our guess is that there is curvature data included in this surface file. Is that so, and is there an easy way to create our own reference surface from a set of subjects?
Thanks in advance,
Janne Hämäläinen
-- Janne Hämäläinen Researcher Helsinki University Central Hospital Department of Radiology janne.hamalainen@hus.fi
1. you'll need the ?h.curv (smoothed mean curvature) and ?h.sulc (convexity, generated by mris_inflate) to morph.
2. The ?h.average.tif are multi-frame tif, with 3 sets of 3 frames. Each set contains mean, variances and dofs for a geometric measure. The 1st set is the curvature of the inflated surface, which is not currently used in the morph. The second set is convexit (sulc) and the third is the smoothed mean curvature.
cheers, Bruce
On Mon, 7 Jun 2004, Hämäläinen Janne wrote:
Hi,
We are planning to use some of our own methodology (mainly segmentation routines) in the Freesurfer pipeline to make cortical thickness analysis between groups. Related to this, there is two things that would need clarification.
- In case we would use our own surfaces for analysis, what surface
properties would be needed by the registration software (mris_register) of Freesurfer, and, more precisely, exactly what curvature files? In the www-site of Graham Wideman, he presents a data flow chart that tells the input and output to the number of programs used in the analysis. He has concluded that the registration would need only the spherical surface file (rh.sphere or lh.sphere), but that can't be correct, right? Some curvature information needs to be used, but what? In the article 'High-Resolution Intersubject Averaging and a Coordinate System for the Cortical Surface' (HBM 1999) it is said that convexity of the surface can be used. But what is the property (and curvature file) used in Freesurfer?
- Then we would of course need our own reference surface for registration,
since we are analysing surfaces constructed differently from the default Freesurfer surfaces. Is there documentation on what data the reference surface (e.g. the rh.average.tif or lh.average.tif) contains? Our guess is that there is curvature data included in this surface file. Is that so, and is there an easy way to create our own reference surface from a set of subjects?
Thanks in advance,
Janne Hämäläinen
-- Janne Hämäläinen Researcher Helsinki University Central Hospital Department of Radiology janne.hamalainen@hus.fi
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Hämäläinen Janne