I'm making progress, but I can't fix this. the output is: start volume math end volume math volume thresholds are nan to nan start raw mapping nan% of in-range, in-mask voxels failed the threshold nan% of in-range, in-mask voxels failed the threshold end raw mapping start correction Out of 240195 cortex nodes: 100% were zero and needed reconstruction 0% were discarded by threshold When reconstructing discarded nodes: 100% were zero and not used 0% were outside threshold and ignored Out of 241248 cortex nodes: 100% were zero and needed reconstruction 0% were discarded by threshold When reconstructing discarded nodes: 100% were zero and not used 0% were outside threshold and ignored end correction
I have: a left and right white, pial, midthickness (average). these surfaces were transformed into caret space much in the way you describe on the wiki, but with an extra bit to resacle to 0.25mm from 1mm. the T1w is perfectly in register with the surfaces. It has 0.25mm voxels, and it's an MTR. I scaled it so that the WM is generally around 110, as in FS. the T2w is a mask of T1w with all non-zero voxels set to 1, so that T1w/T2w=T1w which is what I want since I'm using MTR. the ribbon is the freesurfer ribbon, again reset to 0.25mm voxels. I set the right side of the ribbon to 255 and the left to 120. All the volumes are in perfect register. all have the same AC setting in caret. All have voxel ordering +++. The surfaces are all perfectly registered to the volumes. the thickness metric was obtained by changing the FS max from 5 to 40mm, converting the thickness file as instructed to a metric, dividing that by 4, to get thickness for 0.25mm via the 1mm trick. loading the metric files give plausible values on the surface of about 1mm in V1, up to a maximum of maybe 3mm in dorsal motor regions. The curvature metric also seems exactly right. the topo is derived from the freesurfer wm surface. I've run out of ideas. I'll try making the ribbon a volume of ones. I doubt that'll fix it. sorry about this, it would really be nice to get this myelin map for the monkey. tests using simpler mapping methods indicate it's definitely worth it. We want to see how this relates to some diffusion properties in the GM, and actually tracking termination patterns also. best, Colin
_______________________________________________ caret-users mailing list [email protected] http://brainvis.wustl.edu/mailman/listinfo/caret-users
