many thanks,
the issue is not the size of the voxels. the issue is the dimensions of the volume in units of "voxel", which as I understand it must be 256x256x256 voxels exactly, but hopefully not.
with voxels of any size one just changes the header to 1mm and adjusts calculations later on. that's the easy bit actually. but it doesn't alter how many voxels there are. my data has too many voxels of actual brain MRI. the voxel size in mm isn't really the thing, it's just a number making a claim.
many thanks for the document though. It may not work with monkeys. They are a bit hands on.
Colin
2012/2/25 Jürgen Hänggi j.haenggi@psychologie.uzh.ch
Hi Colin
There is a guide how to force freesurfer using higher resolution than 1x1x1 mm. It was written by Falk Lüsebrink. If you search in the FS archive with:
Measuring cortical thickness with high resolution data
You find some threads on this topic
Cheers Jürgen
On [DATE], "Colin Reveley" <[ADDRESS]> wrote:
Hi -
I would really like to use a volume size greater than 256^3.
I know of all the options and reasons.
what I want is a white matter surface than has fairly evenly spaced nodes that are in approximate match to the voxels they enclose.
the reason is to seed, track and target using FSL surface tractography. the brain is a 250um macaque, and it's 320 voxels long unfortunately. the structural and DWI are the same resolution and in register and acpc aligned. such a shame to waste.
the freesurfer surfaces are imported into CARET and cross-registered to F99. but they keep their mesh if I want.
work with another sample showed that node numbers and even coverage over the WM/GM boundary had a signifiant effect on the FSL metrics, with the original FS WM surface and derivatives yielding the best results over other meshes like F99's or anything made with surefit directly from a downsampled version of the data. that data I could fit into 256^3.
I doubt that the nodes and voxels need to match perfectly, just ballpark and fairly even spacing.
if I used a downsampled volume at 500um to make surfaces in freesurfer, maybe I could upsmaple the surfaces afterwards? but then the pial and white would have to be resampled independently and that's probably no good.
If I have to resample the volume or apply a scaling transform the world won't end. In fact it will take less time to do probably. But it would be nice to use a larger size than 256^3 to take advantage of what I've got.
best,
Colin, Sussex
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Jürgen Hänggi, Ph.D. Division Neuropsychology Institute of Psychology University of Zurich Binzmuehlestrasse 14, PO Box 25 8050 Zurich, Switzerland 0041 44 635 73 97 (phone office) 0041 76 445 86 84 (phone mobile) 0041 44 635 74 09 (fax office) BIN 4.D.04 (office room number) j.haenggi[at]psychologie.uzh.ch (email) http://www.psychologie.uzh.ch/neuropsy/ (website) http://www.juergenhaenggi.ch (private website)
This e-mail (and any attachment/s) contains confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorised copying, disclosure or distribution of the material in this e-mail is strictly forbidden.