There is an option in the new segmentation user interface to produce
"imported" tissue class images, which are suitable for use with DARTEL. I
figured that this would simplify the processing, but it is not properly
documented yet. I'll update the manual when I get a chance.
The cleaner results with the new segmentation are mostly from using a
different set of tissue probability maps. Further work on these would result
in yet cleaner results (another thing on my to do list).
On Tuesday 19 May 2009 17:13, Rick Helmich wrote:
> Dear John Ashburner and list,
> Could anybody tell me how to combine the New Segmentation tool and DARTEL
> for VBM in SPM8?
> I noticed that the "normal" unified segmentation in SPM8 labels some pieces
> of dura as gray matter, but that the New Segmentation tool does a much
> better job.
> However, the normalization parameters that the New Segmentation produces
> (_seg8.mat) are not recognized by DARTEL when I run the "Initial Import"
> It seems to me that the combination of these two tools is at the moment the
> best option for performing VBM analyses, and I would be very interested in
> knowing how to combine them. Or is there another way to get rid of the dura
> from the gray matter maps when using the "normal" unified segmentation?
> Thanks a lot!
> Rick Helmich