Hi All,
as the subject line reveals, I have a question regarding mkanalysis-sess.new (running centos4_x86_64 3.0.1 pub stable, on SuSE 9.3, not that it matters...). My question is, is there any heuristic how to chose the -taumax parameter, that can be used in combination with a brainmask? I am wondering as the example usage from the fsfast workflow specifies a brainmask, but does not set -taumax, is that a useful approach? One more question about the brainmask I have is, (at least without a specified taumax) I still see non-uniform values in the sig volume, while actually expecting close to nothing, but I guess I have not groked how the masking is supposed to work... I would be happy for any pointers about this issue.
And while on the matter, is there a way to allow empty conditions, i.e. use a sparse condition numbering in the .para files (trading the check for missing conditions versus the possibility to use stable identifiers for loads of conditions used over different days).
Ahoi Sebastian
I usually set it to about 20 sec (the interpretation of this parameter is that the temporal autocorrelation will return to 0 beyond this spacing). The brain mask is used to restrict the voxels overwhich the autocorrelation is computed and averaged (don't want to include "air" voxels in the computation). What do you mean that you see non-uniform values in the sig volume?
Sebastian Moeller wrote:
Hi All,
as the subject line reveals, I have a question regarding mkanalysis-sess.new (running centos4_x86_64 3.0.1 pub stable, on SuSE 9.3, not that it matters...). My question is, is there any heuristic how to chose the -taumax parameter, that can be used in combination with a brainmask? I am wondering as the example usage from the fsfast workflow specifies a brainmask, but does not set -taumax, is that a useful approach? One more question about the brainmask I have is, (at least without a specified taumax) I still see non-uniform values in the sig volume, while actually expecting close to nothing, but I guess I have not groked how the masking is supposed to work... I would be happy for any pointers about this issue.
And while on the matter, is there a way to allow empty conditions,i.e. use a sparse condition numbering in the .para files (trading the check for missing conditions versus the possibility to use stable identifiers for loads of conditions used over different days).
Ahoi Sebastian
Hi Doug, hi List,
On 10. Apr 2006, at 20:28 Uhr, Doug Greve wrote:
I usually set it to about 20 sec (the interpretation of this parameter is that the temporal autocorrelation will return to 0 beyond this spacing). The brain mask is used to restrict the voxels overwhich the autocorrelation is computed and averaged (don't want to include "air" voxels in the computation). What do you mean that you see non-uniform values in the sig volume?
Mmh, I was expecting that all the voxels outside the brainmask would not take part in the analysis/calculation at all, or that all the non_brain voxels in the sig_NNN.bfloat volume of the contrasts would be zero as well. Mmh, maybe that is because I did not specify a -taumax. Just started a test run with "-taumax 20", and each "Slice N, NN.NNNN --------------" line is followed by a "NBrainVoxs = NNNN" line, so user error on my part. Redid an analysis right now, and the end result with -taumax 20 has way less significant activation outside the brain proper. Thanks for the explanation.
Ahoi Sebastian
Sebastian Moeller wrote:
Hi All,
as the subject line reveals, I have a question regarding mkanalysis-sess.new (running centos4_x86_64 3.0.1 pub stable, on SuSE 9.3, not that it matters...). My question is, is there any heuristic how to chose the -taumax parameter, that can be used in combination with a brainmask? I am wondering as the example usage from the fsfast workflow specifies a brainmask, but does not set -taumax, is that a useful approach? One more question about the brainmask I have is, (at least without a specified taumax) I still see non-uniform values in the sig volume, while actually expecting close to nothing, but I guess I have not groked how the masking is supposed to work... I would be happy for any pointers about this issue.
And while on the matter, is there a way to allow emptyconditions, i.e. use a sparse condition numbering in the .para files (trading the check for missing conditions versus the possibility to use stable identifiers for loads of conditions used over different days).
Ahoi Sebastian
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
freesurfer@nmr.mgh.harvard.edu
-
Doug Greve -
Sebastian Moeller