Hi Doug,
Sure I can. I bump into problems though, exactly the same problems as mentioned before (http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg14953.html):
Found 136019/147456 (92.2) voxels in mask Creating Design Matrix ... creation time = 0.010 sec DoMCFit = 1 ntptot = 686, nX = 42, DOF = 644 Saving X matrix to /home/action/fraleo/Data/Exp1_Saccades_phase/S1_FS/bold/rtopy.S1_FS.rh/Xtmp.mat ??? Error using ==> svd Input to SVD must not contain NaN or Inf.
Error in ==> cond at 40 s = svd(A);
Error in ==> fast_selxavg3 at 247 XCond = cond(XtX);
Any ideas why this is/how I can solve this? Same result for multiple subjects, not sure where it comes from.
Thanks a lot in advance,
kind regards,
Frank ----- Original Message ----- From: "Douglas N Greve" greve@nmr.mgh.harvard.edu To: "Frank Leoné" f.leone@donders.ru.nl Cc: freesurfer@nmr.mgh.harvard.edu Sent: Monday, October 18, 2010 5:08:42 PM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna Subject: Re: [Freesurfer] Retinotopic mapping of only polar data]
Can you let me know when you have processed a set with the new tools? It's going to be a lot easier for me to instruct you on how to do those things.
doug
f.leone@donders.ru.nl wrote:
Dear Doug,
Thanks a lot for the reply, that really helps. I see that the new function is more integrated with the rest of the standard analyses, so it now also wants to do motion correction ed. it seems, so didn't run the full thing yet. Still, if I make the pictures using the new functions, but an old h.nii, I get the same kind of pictures, so can you maybe also help me with the interpretation questions:
- How to interpret the color codes / values on the inflatedsurface?
- How to threshold them on significance?
(Third one: how to interpolate them to get away from the patchy representation?)
You suggest to smooth the data btw, thought that wasn't standard procedure when doing retinotopic mapping, at least not with a fwhm of 5.
Thanks again,
kind regards, Frank----- Oorspronkelijk bericht ----- Van: "Douglas N Greve" greve@nmr.mgh.harvard.edu Aan: "Frank Leoné" f.leone@donders.ru.nl Cc: freesurfer@nmr.mgh.harvard.edu Verzonden: Donderdag 14 oktober 2010 22:52:21 GMT +01:00 Amsterdam / Berlijn / Bern / Rome / Stockholm / Wenen Onderwerp: Re: [Freesurfer] Retinotopic mapping of only polar data]
Hi Frank, sfa-sess does not exist in version 5 of FreeSurfer because its functions have been incorporated into selxavg3-sess. It will be a lot easier to support you if you move to version 5. Unfortunately, I have been slow in getting it documented. Having said that, I think the new version is a lot easier to use. It is very similar to regular task studies, which is pretty well documented (see http://nmr.mgh.harvard.edu/~greve/fsfast.intro.ppt).
For retinotopy, there are only a few deviations: 1.a. When you create the analysis, specify something like: mkanalysis-sess -a rtopy.self.lh -surface self lh -TR 2 -fwhm 5 -retinotopy 30 -paradigm rtopy.par where "-retinotopy 30" means that it is a retinotopy study with a period of 30 sec 1.b Run selxavg3-sess (instead of sfa-sess) 1.c Run tksurfer-sess, like tksurfer-sess -s sess -a rtopy.self.lh This will show you the eccen and polor contrasts 2.a Create a field sign (with an occipital patch) with: fieldsign-sess -a rtopy.lh -s sess -occip 2.b View the fieldsign tksurfer-sess -s sess -a rtopy.self.lh -fieldsign
The above assumes that you want to analyze the data on the subject's left hemi. Make and run anther analysis for the right hemi
doug
f.leone@donders.ru.nl wrote:
Hi everyone,
The e-mail included below I sent some time ago, maybe it slipped through. In the mean time I did find out I can plot the results with tksurfer-sess like
tksurfer-sess -s S2_FS -h rh -analysis rtopy -c polar -map h
But then two big questions remain:
- How to interpret these color codes?
- How to threshold them on significance?
(Third one: how to interpolate them to get away from the patchy representation?)
The other questions stated below are also still relevant.
Asked this before, but can we add stuff ourselves to the wiki? Found out quite a bit about the retinotopic inner workings which I wasn't able to find anywhere else, might help some others.
kind regards, FrankOld topic, but it is directly related, so I'll just continue it here.
In the end I have the spots on the right place, but now it's not clear to me how to interpret the map. I have created three overlays (sign, log10 * sin, log10 * cos, like mentioned in selfreq.pdf) and loaded them all three in tksurfer. Most important question now: how to relate these three to get one figure instead of three separate?
Other more minor questions:
- What do the values actually mean? Apparently I can threshold on the
basis of them, but that does't make sense to do, because the direction info is also in there, right? I should probably threshold the F-values first, before overlaying them on the inflated brain.
- What happened to sfa-sess in the new version of freesurfer? I can't
seem to find it.
- Why does the sfa-sess script check whether amp data is present?
Looking at the script, the line that checks whether there is amplitude data can safely be removed.
- Just like the original Sereno paper, I have 20s pre and post stimulus
presentation time. I now only use the data where the convolved DM predicts there will be signal, so discarding the first 20s and, due to the HRF, keeping almost the entire second 20s. Is this the correct thing to do, or should I rather keep everything / make sure I have an exact number of periods for each stimulus?
Thanks a lot in advance,
kind regards, FrankOn Mon, 2010-01-18 at 17:32 -0500, Douglas N Greve wrote:
If the activation is showing up in the wrong places, then that probably means that the registration is off. Have you checked the registration.
doug
f.leone@donders.ru.nl wrote:
Probably my e-mail just slipped through, so just a short notice that I still have the problem with the overlay mentioned in the last e-mail:
"I now have a polar map overlayed on my inflated brain. But something seems to be amiss, because it plots it mainly on the temporal lobe, with nothing on the most dorsal slices, while actually I only scanned the 18 most dorsal (oriented along the occipital-parietal axis), sot here shouldn't be anything in the temporal or inferior frontal lobe. How can this be/where is it going wrong?
Moreover, I'm not sure on the format used by the polar map: the scale bar states it ranges from 1.18 to 1.18, but I don't see the sense in that. What will come out? Radians I pressume?"
Is it possible to edit the wikipages ourselves btw? I could add some more / update the info on the retinotopic mapping to help other people when I got it all figured out.
kind regards, thanks in advance,
Frank----- Original Message ----- From: "f.leone@donders.ru.nl" f.leone@fcdonders.ru.nl To: freesurfer@nmr.mgh.harvard.edu Sent: Monday, January 11, 2010 9:15:43 AM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna Subject: Re: [Freesurfer] Retinotopic mapping of only polar data
Hi everyone,
Doug: thanks a lot for the reply, that indeed worked, I now have a polar map overlayed on my inflated brain. But something seems to be amiss, because it plots it mainly on the temporal lobe, with nothing on the most dorsal slices, while actually I only scanned the 18 most dorsal (oriented along the occipital-parietal axis), sot here shouldn't be anything in the temporal or inferior frontal lobe. How can this be/where is it going wrong?
Moreover, I'm not sure on the format used by the polar map: the scale bar states it ranges from 1.18 to 1.18, but I don't see the sense in that. What will come out? Radians I pressume?
kind regards, thanks a lot in advance, Frank----- Original Message ----- From: "Douglas N Greve" greve@nmr.mgh.harvard.edu To: "Frank Leoné" f.leone@donders.ru.nl Cc: freesurfer@nmr.mgh.harvard.edu Sent: Friday, January 8, 2010 4:26:21 PM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna Subject: Re: [Freesurfer] Retinotopic mapping of only polar data
You can generate a register.dat with either fslregister-sess or spmregister-sess, or, if you have it in your version, use bbregister (note that there is not a -sess version for this yet).
You can then use mri_vol2surf to resample the h.nii file to the surface, then view it in tksurfer. You can also extract certain frames with mri_convert with the --frame option.
doug
f.leone@donders.ru.nl wrote:
Hi everyone,
Still not working fully, partly because I'm missing some info on the file formats. What I now did was a run of selfreqavg with four input polar sessions, which ran succesfully. This gives me the expected h.nii. Now I either want to
- use this h file directly, using paint-sess (get message register.dat is missing) and sliceview-sess (get message it can't determine format h-offset, which makes sense, because it doesn't exist)
- or extract w files from the first three layers of the nii files and superimpose them in tksurfer. I don't know how I should extract them though. Can use the SPM surfrend functions, but I can't imagine that is the way, or is it?
Hopefully someone can help me along either of the two paths, I'm really eager to see whether the retinotopic analysis worked. Thanks a lot in advance!
kind regards, Frank----- Original Message ----- From: "f.leone@donders.ru.nl" f.leone@fcdonders.ru.nl Cc: freesurfer@nmr.mgh.harvard.edu Sent: Monday, January 4, 2010 9:35:51 AM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna Subject: Re: [Freesurfer] Retinotopic mapping of only polar data
Hi everyone,
I fixed it with the help of Jascha: when using the full path, e.g. /home/frank/etc.../007/f.nii it does find them. So it is now running. Thanks for the help!
kind regards, Frank----- Original Message ----- From: "Douglas N Greve" greve@nmr.mgh.harvard.edu To: "Frank Leoné" f.leone@donders.ru.nl Cc: freesurfer@nmr.mgh.harvard.edu Sent: Tuesday, December 22, 2009 7:58:23 PM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna Subject: Re: [Freesurfer] Retinotopic mapping of only polar data
Do you have read permissions to the file? If so, can you send the log file created by sfa-sess?
f.leone@donders.ru.nl wrote:
Thanks a lot for the reply. I now bump into different problems though: it doesn't understand my f.nii file, see below. I run:
frank@frank-desktop:~/PhD/data/Exp1_Saccades_phase$ sfa-sess -a rtopy -s FSFAST_S1/
and I think the main error line is:
ERROR: cannot determine format of 007/f
>> ??? Attempt to reference field of non-structure array. >> >> >> >> >> 007 is my first session (is a dir) and in there is one f.nii, which includes all the scans of that run. . Previously I tried it with separate scan files, but that didn't work either. So what does it expect? And will it work with data preprocessed in SPM instead in FreeSurfer? The rest of the dir structure is setup as specified in the guide
Hopefully someone can help me, thanks a lot in advance,
kind regards, Frankfrank@frank-desktop:~/PhD/data/Exp1_Saccades_phase$ mkanalysis-sess.new -force -a rtopy -TR 2.12 -designtype retinotopy -paradigm rtopy.par -funcstem f -ncycles 8 INFO: analysis rtopy exists, but overwrite forced by user. Completed successfully frank@frank-desktop:~/PhD/data/Exp1_Saccades_phase$ sfa-sess -a rtopy -s FSFAST_S1/
sfa-sess logfile is /home/frank/PhD/data/Exp1_Saccades_phase/log/sfa-sess-bold-rtopy-0912221252.log
/home/frank/PhD/data/Exp1_Saccades_phase/FSFAST_S1 Tue Dec 22 12:52:30 CET 2009 INFO (FSFAST_S1): RunList = 007 008 009 010 011
/home/frank/PhD/data/Exp1_Saccades_phase/FSFAST_S1/bold selfreqavg -TR 2.12 -o rtopy/polar/h -i 007/f -i 008/f -i 009/f -i 010/f -cfg /home/frank/PhD/data/Exp1_Saccades_phase/rtopy/analysis.cfg -parname rtopy.par
--- Parsing Config File: /home/frank/PhD/data/Exp1_Saccades_phase/rtopy/analysis.cfg ---- -ncycles 8 -delay 0 -nskip 0 -fwhm 0 DirectionList: +1 +1 +1 -1 Warning: Unable to open display 'iconic'. You will not be able to display graphics on the screen.
< M A T L A B (R) > Copyright 1984-2009 The MathWorks, Inc. Version 7.8.0.347 (R2009a) 64-bit (glnxa64) February 12, 2009To get started, type one of these: helpwin, helpdesk, or demo. For product information, visit www.mathworks.com.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Extension format = nii >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>> hanrad = 0 >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ERROR: cannot determine format of 007/f >>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >> ??? Attempt to reference field of non-structure array. >> >> >> >> >>
>> ??? Attempt to reference field of non-structure array. >> >> >> >> >>
>> ??? Attempt to reference field of non-structure array. >> >> >> >> >>
>>>> ??? Attempt to reference field of non-structure array. >>>> >>>> >>>> >>>> >>>>
>>>> ??? Undefined function or variable 'lastslice'. >>>> >>>> >>>> >>>> >>>>
>> ??? Attempt to reference field of non-structure array. >> >> >> >> >>
>>>>>> ??? Attempt to reference field of non-structure array. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ??? Index exceeds matrix dimensions. >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>>>>>> ??? Attempted to access tmp(NaN); index must be a positive integer or logical. >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>>
>> ??? Attempted to access tmp(NaN); index must be a positive integer or logical. >> >> >> >> >>
>> ??? Attempted to access tmp(NaN); index must be a positive integer or logical. >> >> >> >> >>
>> ??? Attempted to access tmp(NaN); index must be a positive integer or logical. >> >> >> >> >>
>> ??? Attempted to access tmp(NaN); index must be a positive integer or logical. >> >> >> >> >>
>> ??? Attempted to access tmp(NaN); index must be a positive integer or logical. >> >> >> >> >>
>>>>>>>>>>>>>>>>>>>>>>>>>> Slice >> ??? Undefined function or variable 'lastslice'. >>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>>>>>>>>>>>
>>>>>> ??? Attempt to reference field of non-structure array. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> Error in ==> MRIwrite at 94 hdr.dim = [mri.volsize(2) mri.volsize(1) mri.volsize(3) mri.nframes];
>>>>>> ??? Attempt to reference field of non-structure array. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> Error in ==> MRIwrite at 94 hdr.dim = [mri.volsize(2) mri.volsize(1) mri.volsize(3) mri.nframes];
>>>>>> ??? Attempt to reference field of non-structure array. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> Error in ==> MRIwrite at 94 hdr.dim = [mri.volsize(2) mri.volsize(1) mri.volsize(3) mri.nframes];
>>>>>>>> Saving header to rtopy/polar/h >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >> Could not open rtopy/polar/h.sfa for writing >> >> >> >> >> quiting matlab selfreqavg COMPLETED mri_convert rtopy/polar/h.nii -o rtopy/polar/h-offset.nii --frame 10 mri_convert rtopy/polar/h.nii -o rtopy/polar/h-offset.nii --frame 10 niiRead(): error opening file rtopy/polar/h.nii
----- Original Message ----- From: "Douglas N Greve" greve@nmr.mgh.harvard.edu To: "Frank Leoné" f.leone@donders.ru.nl Cc: freesurfer@nmr.mgh.harvard.edu Sent: Wednesday, December 16, 2009 5:12:41 PM GMT +01:00 Amsterdam / Berlin / Bern / Rome / Stockholm / Vienna Subject: Re: [Freesurfer] Retinotopic mapping of only polar data
If you run the retinotopy analysis, it will expect eccen. If it stops without doing the polar analysis, you can "fool" it by copying the polar data into the eccen directory.
doug
f.leone@donders.ru.nl wrote:
> Dear all, > > I just started using FreeSurfer, specifically for retinotopic data, on the moment only polar data. So I followed the guide as mentioned on http://surfer.nmr.mgh.harvard.edu/fswiki/FsFastIndividualRetinotopyAnalysis . This does seem to work to some extend, except that it complains that I don't have eccentricity data. He is right about that, but is the analysis also doable with the eccentricity data? I really hope so, thanks a lot in advance, > > kind regards, > > Frank > > > > > > >
----- Oorspronkelijk bericht ----- Van: "Frank Leone" f.leone@donders.ru.nl Aan: "f leone" f.leone@donders.ru.nl Verzonden: Woensdag 6 oktober 2010 17:16:43 GMT +01:00 Amsterdam / Berlijn / Bern / Rome / Stockholm / Wenen Onderwerp: [Fwd: Re: [Freesurfer] Retinotopic mapping of only polar data]
freesurfer@nmr.mgh.harvard.edu
-
f.leone＠donders.ru.nl