Re: [Freesurfer] Registration & segmentation with FSL - obtaining correlation matrix..

The cost (from .mincost) looks fine. Can you send a pic of the registration inaccuracy? doug On 02/25/2013 11:53 AM, Sudhin A. Shah wrote:

Hi Doug,

Is the BBR cost function in anat2exf.register.dat.bbr.init? mwtp5 3.750000 4.000000 0.150000 9.689088e-01 6.952747e-02 -2.374479e-01 6.737892e+00 2.456762e-01 -1.566831e-01 9.566056e-01 -2.835866e+01 2.930631e-02 -9.851995e-01 -1.688930e-01 1.765446e+01 0 0 0 1 round

Or in anat2exf.register.dat.mincost 0.545185 785.020257 804.709380 1.199352

This looks ok under manual inspection - although the head was not in an ideal position (patient data). However, I cannot extract data from ROI label 2013.

Thanks,

---

From: Douglas N Greve [greve@nmr.mgh.harvard.edu] Sent: Monday, February 25, 2013 11:39 AM To: Sudhin A. Shah; Freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Registration & segmentation with FSL - obtaining correlation matrix..

First check the value of the BBR cost function. If it is over .8, then something is wrong. To check whether it is an LR flip, you can flip the volume LR and re-run the registration to see if it gets better. doug

On 02/25/2013 11:19 AM, Sudhin A. Shah wrote:

...

Hi Doug,

When I initially pulled out the correlation values (using the matlab code below) after the automated registration process: regfeat2anat & aseg2feat, some ROIS reported NaNs.

When I manually checked the registration and adjusted it using tkregister2, I found myself stretching/scaling to make it fit. Consequently I do not have NaNs BUT I am now realising that I should'nt have to scale.

1. How can I check if left-right reversals have happened - its not obvious by eye
2. I am going to re-run the automated registration and I can send you a pic and the cost function..

Thanks, S

---

From: Douglas N Greve [greve@nmr.mgh.harvard.edu] Sent: Monday, February 25, 2013 10:44 AM To: Bruce Fischl Cc: Sudhin A. Shah Subject: Re: [Freesurfer] Registration & segmentation with FSL - obtaining correlation matrix..

Hi Sudhin, I'm not sure what the problem is. NaNs in the data? Stretching the registration? You should not need to stretch the reg.Are you 100% positive that the func and anat come from the same subject and there have been on left-right reversals? What is the final value of the cost function? doug

On 02/25/2013 10:06 AM, Bruce Fischl wrote:

...

Hi Sudhin

Doug would have some. You definitely shouldn't have to stretch/scale to get it to fit. If you do, something critical is wrong. Also, please cc the list so that the right person (Doug in this case!) can answer.

cheers Bruce

On Mon, 25 Feb 2013, Sudhin A. Shah wrote:

...

Hi Bruce,

They were a result of the registration (between structural and functional) being 'off'. When I manually correct this using tkregister2, they were fine. However, I found that I had to stretch (scale) to get them to fit so (i started a new thread regarding this) are there image examples on what this should look like so I can use as a guide?

Thanks, S

On Feb 25, 2013, at 9:27 AM, Bruce Fischl wrote:

...

Hi Sudhin

are you sure that the NaNs don't exist before registration and extraction? Cany you check? I don't see why anything we would have done would introduce them.

cheers Bruce On Fri, 22 Feb 2013, Sudhin A. Shah wrote:

...

         Hello

I have run free surfer and co-registered with the rsfMRI via FSL (reg-feat2anat & aseg2feat) When I extract the values (as per the matlab code below), I see NaNs in certain regions. Of the 19 datasets that I have analyzed I have see NaNs for the following labels 1006 1033 2006 2013 2033 Is this a known problem? Is there a solution? Thanks, S On Jan 31, 2013, at 3:41 PM, Douglas N Greve wrote:

   Sorry, I don't think we have such a program. You could do it

in matlab fairly easily, eg,

   f = MRIread('filtered_func_data');
   fmat = fast_vol2mat(f);
   a = MRIread('aparc+aseg');
   seglist = unique(a.vol(:));
   seglist = seglist(2:end); % remove segid=0 (unkown)
   clear roimean
   for nthseg = 1:length(seglist)
     ind = find(a.vol == seglist(nthseg));
     roimean(:,nthseg) = mean(fmat(:,ind),2);
   end

   m = roimean'*roimean;

   On 01/31/2013 01:29 PM, Sudhin A. Shah wrote:

         Hello,

         This worked perfectly 2 years ago :), but now I am

having some trouble.

         I run reg-feat2anat & aseg2feat with no problem.

         I now need a correlation matrix of every ROI (created by

freesurfer)

         against every other ROI. For this I used @ROI_Corr_Mat

(http://afni.nimh.nih.gov/pub/dist/doc/program_help/@ROI_Corr_Mat.html).

         Again - this worked fine in the past.

         Now I get an error (working with AFNI to see if it can

be fixed).

         Question: Is there any alternative to getting this

correlation matrix?

         i.e every ROI against every other ROI?

         Thanks,

         S

         _______________________________________________

         Freesurfer mailing list

         Freesurfer@nmr.mgh.harvard.edu

https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

   --
   Douglas N. Greve, Ph.D.
   MGH-NMR Center
   greve@nmr.mgh.harvard.edu
   Phone Number: 617-724-2358
   Fax: 617-726-7422

   Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
   FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
   Outgoing:

ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/

   _______________________________________________
   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. _______________________________________________ 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. _______________________________________________ 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.

-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422

Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/

-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422

Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/