 |
 |
Hi Ged, I was running the script on ubuntu, but shell is set to bash (not dash). The error it gave was "cannot find concat matrix" which made me think there was something wrong with the script. It did that no matter what I set for the s/qform settings at the end. It did output a matrix, but this matrix was not correct reduced the data in size drastically, and put it in the wrong place. I would be happy to test your version of the script if you send it to me personally, rather than through the list so it doesn't get blocked. Perhaps I made some mistake when converting the script from avw to fsl (I did it with a find and replace). Peace, Matt. -----Original Message----- From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of DRC SPM Sent: Wednesday, October 17, 2007 8:50 AM To: [log in to unmask] Subject: Re: [FSL] MNI Space in MRIcro and FSL Sorry, jiscmail rejected the attachment for some reason... On 17/10/2007, DRC SPM <[log in to unmask]> wrote: > Hi Matt, > > I wonder why it didn't work? I've just run it through avw2fsl (and > also changed /bin/sh to /bin/bash in the #! line for good measure, as > I seem to recall Debian/Ubuntu was unhappy with sh), and tested it on > a pair of images with different image dimensions and different q- and > s-forms (from co-registration in SPM5) as follows, and the result > looks fine. > > flirt_qsinit.sh src ref src2ref.mat > flirt -in src -ref ref -out src2ref -init src2ref.mat -applyxfm > fslview ref src2ref > > I've attached my copy in case it helps, if it doesn't, hopefully Mark > or Steve could take a look at your data if you upload it? (Or I could, > if you put it somewhere I can see it). > > Best, > Ged > > > On 17/10/2007, Matt Glasser <[log in to unmask]> wrote: > > Thanks Ged. > > > > I couldn't get the script to work, even when I changed avw to fsl, but the > > SPM method worked fine. > > > > Peace, > > > > Matt. > > > > -----Original Message----- > > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf > > Of Ged Ridgway > > Sent: Tuesday, October 16, 2007 10:21 AM > > To: [log in to unmask] > > Subject: Re: [FSL] MNI Space in MRIcro and FSL > > > > Hi Matt, > > > > Mark J posted a script that can use the NIfTI header info to generate > > a FLIRT transformation: > > http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0701&L=FSL&P=R20004 > > I think if you apply this transformation with flirt, with the standard > > as the -ref, then your FSLview results should match your MRIcron ones. > > > > I'm guessing MRIcron resamples a binary ROI in the space of the > > standard image, with either nearest neighbour interpolation or linear > > followed by re-binarisation at a threshold of 0.5 -- you might need to > > find out its implementation details to exactly reproduce the results. > > Possibly MRIcron has an option to write out the resampled ROI, that it > > presumably stores internally for display purposes? > > > > Hope that works, > > Ged. > > > > P.S. Another option, if you use SPM (sorry to mention the enemy on > > this list again!), is to use "coreg"->"reslice" with the standard > > image defining the space and the ROI in the "other images" (or words > > to that effect). The resulting ROI images prefixed with "r" should > > then overlay properly in FSLview. Check out spm_reslice if you want to > > avoid using the GUI. > > > > > > Matt Glasser wrote: > > > Hi Steve, > > > > > > How do you suggest I go about correcting for this difference? I could > > just > > > do it with nudge, but I would prefer an automated way that does not have > > the > > > chance of human error. > > > > > > Thanks, > > > > > > Matt. > > > > > > -----Original Message----- > > > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf > > > Of Steve Smith > > > Sent: Tuesday, October 16, 2007 8:27 AM > > > To: [log in to unmask] > > > Subject: Re: [FSL] MNI Space in MRIcro and FSL > > > > > > Hi, > > > > > > On 16 Oct 2007, at 12:58, Matt Glasser wrote: > > > > > >> Hi Steve, > > >> > > >> Well okay, but why would the ROIs line up with FSL's standard brain in > > >> MRIcro, but not in FSL itself? I used the same standard space > > >> image in both > > >> programs > > > > > > Ah - I hadn't appreciated that - sorry... > > > > > >> and it was only misaligned in one of them, so I don't know how I > > >> could use a flirt transform in this case. > > > > > > Right - in that case the difference presumably is because MRICRO uses > > > the co-ordinate centre from the header in determining display > > > centering, whereas FSLView only uses the raw voxels for working out > > > display (the co-ordinate centre is only used in the reporting of the > > > mm co-ordinates, not in the image display). > > > > > > Cheers, Steve. > > > > > > > > > > > >> Thanks, > > >> > > >> Matt. > > >> > > >> -----Original Message----- > > >> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On > > >> Behalf > > >> Of Steve Smith > > >> Sent: Tuesday, October 16, 2007 2:00 AM > > >> To: [log in to unmask] > > >> Subject: Re: [FSL] MNI Space in MRIcro and FSL > > >> > > >> HI - I presume this is because of a change in field-of-view in the > > >> different standard space images. One easy way to solve that is to use > > >> flirt to estimate the transform between the standard space image in > > >> SPM/mricro and the FSL standard space image, and use that instead of > > >> eye.mat in your call to flirt resample the ROI. > > >> > > >> Cheers, Steve. > > >> > > >> > > >> On 16 Oct 2007, at 00:17, Matt Glasser wrote: > > >> > > >>> Hi, > > >>> > > >>> > > >>> > > >>> I have a set of ROIs from SPM in MNI space that I want to load into > > >>> FSL. When I take the standard FSL brain MNI152_T1_1mm_brain and > > >>> load it in MRIcro and overlay the ROI, it looks fine, like the > > >>> first screenshot: http://www.ma-tea.com/~Matt/MRIcroOnMNI.png > > >>> However, if I resample the 4X4X4mm ROI so it can be overlayed on > > >>> this brain in FSLView, using: > > >>> > > >>> > > >>> > > >>> flirt -in 4X4X4mmROI -ref MNI152_T1_1mm_brain -applyxfm -init > > >>> eye.mat -out 1X1X1mmROI where eye.mat is the identity matrix: > > >>> > > >>> > > >>> > > >>> 1 0 0 0 > > >>> > > >>> 0 1 0 0 > > >>> > > >>> 0 0 1 0 > > >>> > > >>> 0 0 0 1 > > >>> > > >>> > > >>> > > >>> The result is incorrect, shown by the second screenshot in FSLView: > > >>> http://www.ma-tea.com/~Matt/FSLOnMNI.png > > >>> > > >>> > > >>> > > >>> The ROI file header is the following: > > >>> > > >>> > > >>> > > >>> <nifti_image > > >>> > > >>> nifti_type = 'ANALYZE-7.5' > > >>> > > >>> image_offset = '0' > > >>> > > >>> ndim = '4' > > >>> > > >>> nx = '41' > > >>> > > >>> ny = '48' > > >>> > > >>> nz = '35' > > >>> > > >>> nt = '1' > > >>> > > >>> dx = '4' > > >>> > > >>> dy = '4' > > >>> > > >>> dz = '4' > > >>> > > >>> dt = '1' > > >>> > > >>> datatype = '16' > > >>> > > >>> nbyper = '4' > > >>> > > >>> byteorder = 'LSB_FIRST' > > >>> > > >>> cal_min = '0' > > >>> > > >>> cal_max = '1' > > >>> > > >>> descrip = 'FSL4.0' > > >>> > > >>> sform_code = '2' > > >>> > > >>> sto_xyz_matrix = '-4 0 0 80 0 4 0 -112 0 0 4 -52 0 0 0 1' > > >>> > > >>> num_ext = '0' > > >>> > > >>> /> > > >>> > > >>> > > >>> > > >>> I would appreciate any assistance with this as I am unfamiliar with > > >>> this type of issue if it has to do with the analyze file format or > > >>> header. > > >>> > > >>> > > >>> > > >>> Thanks, > > >>> > > >>> > > >>> > > >>> Matt. > > >>> > > >>> > > >> > > >> ---------------------------------------------------------------------- > > >> -- > > >> --- > > >> Stephen M. Smith, Professor of Biomedical Engineering > > >> Associate Director, Oxford University FMRIB Centre > > >> > > >> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > > >> +44 (0) 1865 222726 (fax 222717) > > >> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > > >> ---------------------------------------------------------------------- > > >> -- > > >> --- > > > > > > > > > ------------------------------------------------------------------------ > > > --- > > > Stephen M. Smith, Professor of Biomedical Engineering > > > Associate Director, Oxford University FMRIB Centre > > > > > > FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK > > > +44 (0) 1865 222726 (fax 222717) > > > [log in to unmask] http://www.fmrib.ox.ac.uk/~steve > > > ------------------------------------------------------------------------ > > > --- > > > > > > >