[Freesurfer] request for guide (fwd)
I am forwarding Hengameh's questions to the list since I am not sure about the answers and somehow she couldn't send emails to the list (do people need to subscribe to the list first?). Please help if you know the answers to her questions. Thanks, -Xiao From: hengameh haghighat <[EMAIL PROTECTED]> I have three problems with FreeSurfer software. 1- I don't know how to segment the hippocampus with FreeSurfer software. For this reason i segmentated the hippocampus with another software (MRIcro) and then gave these slides to FreeSurfer software for measuring volume of hippocampus. The obtained volume was greater than the normal volume. For this reason I made a phantom with the specific volume of 27.142 cm3 to obtained correction factor. 2- In the guide's FreeSurfer it is said when watershed algorithm is corrupted , change the preflooding height to 30% in the setting section of RmSkull tab in expert preference , however I didn't get any result. 3- when an analyze file including one slide or all of the taken slides are given to the FreeSurfer software the watershed algorithm is corrupted and the volume is not calculated . 4- How much is the difference between the volume of the subject brain with FreeSurfer software (Albert) and the obtained volume in FreeSurfer software? how have you obtained it's correction factor? I would be thankful if you help me in this matter. I look forward to hearing for you soon Sincerely Yours ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] labels
Could you pls forward the recon-all.log file to me? mris_ca_label should be part of -stage4b. -Xiao On Mon, 14 Nov 2005, Sasha Wolosin wrote: I looked through the recon-all.log file, but I could not find any references to mris_ca_label or mris_anatomical_stats, do you know what else I can search for? I found this warning in several places, but I don't know if it is related: INFO: Volume /tmp/mritotal_12360/orig_8_dxyz.mnc cannot be found. In the recon-all.status file there is another warning: Euler: NHoles = WARNING: Number of holes is not zero This subject was previously manually edited with an older version (oct 2003 release). I made four copies of the subject and ran each of the following in both the 2003 release and the latest stable release: recon-all -subjid subject -stage3 -stage4a -stage4b recon-all -subjid subject -all I only saw the warnings in subjects run with the stable release, (in the 2003 release NHoles =0). Could I be missing some files? Thanks, Sasha Xiao Han <[EMAIL PROTECTED]> >>> If you run "recon-all -all ...", then mris_ca_label and mris_anatomical_stats should be included as the final two steps. You can check the recon-all.log file under each subject's script directory to see why mris_ca_label wasn't executed in your previous run of recon-all. -Xiao On Mon, 14 Nov 2005, Sasha Wolosin wrote: That might have been what happened. I loaded the annotation to the other surfaces, but they also looked strange. Then I re-ran mris_ca_label and the labels look normal. So are mris_ca_label and mris_anatomical_stats not part of the recon-all script (I am using the stable release)? Thanks, Sasha Xiao Han <[EMAIL PROTECTED]> >>> hmm, that's wierd. Did you recomputed the white surface after the annotation file was generated? You can try to load the annotation to other surfaces of the same hemisphere, like ?h.sphere, ?h.sphere.reg, to see whether the same scattered appearance still exists. -Xiao On Mon, 14 Nov 2005, Sasha Wolosin wrote: I double-checked, the correct hemisphere is loaded Xiao Han <[EMAIL PROTECTED]> >>> It's not normal. It's very likely that you load a rh annotation to a lh surface or the other way around. Pls double check. -Xiao On Mon, 14 Nov 2005, Sasha Wolosin wrote: Dear all, When I load the ?h.aparc.annot with the ?h.white surface in tksurfer, the labels look very scattered, is this normal? (please see attached jpeg) Thanks, Sasha Disclaimer: The materials in this e-mail are private and may contain Protected Health Information. Please note that e-mail is not necessarily confidential or secure. Your use of e-mail constitutes your acknowledgment of these confidentiality and security limitations. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying, distribution, or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this e-mail in error, please immediately notify the sender via telephone or return e-mail. ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] labels
It's not normal. It's very likely that you load a rh annotation to a lh surface or the other way around. Pls double check. -Xiao On Mon, 14 Nov 2005, Sasha Wolosin wrote: Dear all, When I load the ?h.aparc.annot with the ?h.white surface in tksurfer, the labels look very scattered, is this normal? (please see attached jpeg) Thanks, Sasha Disclaimer: The materials in this e-mail are private and may contain Protected Health Information. Please note that e-mail is not necessarily confidential or secure. Your use of e-mail constitutes your acknowledgment of these confidentiality and security limitations. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying, distribution, or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this e-mail in error, please immediately notify the sender via telephone or return e-mail. ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] automated parcellation
Hi Goulven, There should be a recon-all.log file under the $SUBJECTS_DIR/subject/script/ directory, if you processed the data using the "recon-all" script. The log files you attached were produced by the tools to visuallize the results, not by the programs that create the label result. Thus we couldn't tell whether anything went wrong during the processing stage from them. -Xiao On Thu, 3 Nov 2005, Goulven Josse wrote: We found the same result, in the same STS region, with different versions : the latest version in january of 2005 for Mac OSX 10.3 and the latest version (installed last week) for Linux / RedHat (again, we don't find this in all subjects). We don't have a recon-all.log file, but i attached other log files coming from a subject processed with the version installed on the Linux / RedHat version. If necessary, we could re-run a subject in order to get a recon-all.log file. goulven ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] coregistration
You may try FLIRT from FSL. http://www.fmrib.ox.ac.uk/fsl/flirt/ -Xiao On Fri, 19 Aug 2005 [EMAIL PROTECTED] wrote: I always have a problem to coregistrate T1 images to Talairach, Is there any way to improve the accuracy of coregistration besides tkregister2? Thanks, Antao ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] mri_watershed
Hi Sasha, If you use tkmedit to visualize the results, then the contrast change you saw may be just due to the display problem (not exactly a problem) of tkmedit. You can move the mouse around the image to actually see whether the image intensity values were changed or not. I used mri_watershed a lot, and never saw it changed the original image intensity (in the brain region). But tkmedit will adjust its display contrast according to the dynamic range of the original image. So once skull is removed, the max and min intensity values of the image are very likely to be different, and the image will then look differently in tkmedit. -Xiao On Fri, 19 Aug 2005, florent segonne wrote: Dear Sasha, mri_watershed was initially designed to skull strip orig images, without any preprocessing step, such as intensity normalization or contrast modification. So, This certainly seems like a reasonable thing to do. However, mri_watershed might normalize intensities during the process but should NOT generate skull-stripped images with different intensities. We'll quickly look into this and let you know if this is a bug that needs to be fixed. Cheers, Florent On Fri, 19 Aug 2005, Sasha Wolosin wrote: We use an anatomical analysis program called Brain Image in our lab. We would like to speed up our analysis in Brain Image by replacing manual skull stripping with FreeSurfer's automated skull strip program. For this we would need to do a skull strip alone, without adjusting intensity values or contrast. Is this possible? Is this a reasonable thing to do? I have run mri_watershed with the input from orig on several subjects, and the output volume looks much brighter than the orig volume. Does mri_watershed normalize intensities or change the contrast in order to strip the skull? Thanks, Sasha Sasha Wolosin Research Assistant Developmental Cognitive Neurology Kennedy Krieger Institute 707 N. Broadway Baltimore, MD 21205 ph: (443) 923-9270 Disclaimer: The materials in this e-mail are private and may contain Protected Health Information. Please note that e-mail is not necessarily confidential or secure. Your use of e-mail constitutes your acknowledgment of these confidentiality and security limitations. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying, distribution, or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this e-mail in error, please immediately notify the sender via telephone or return e-mail. ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] skull strip
Yes, skull-stripping is pretty robust. You can definitely try that on your data. -Xiao On Wed, 10 Aug 2005, Sasha Wolosin wrote: Is it possible to do skull stripping without motion correction or intensity normalization? Disclaimer: The materials in this e-mail are private and may contain Protected Health Information. Please note that e-mail is not necessarily confidential or secure. Your use of e-mail constitutes your acknowledgment of these confidentiality and security limitations. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying, distribution, or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this e-mail in error, please immediately notify the sender via telephone or return e-mail. ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] tkregister2 help
The final objective function seems too high, which indicates that mritotal wasn't able to find a good registration. One approach worth trying is to do the skull-stripping first, then register the skull-stripped brain to the talairach space. From my experience, it can correct the talairach registration problem in most cases. -Xiao On Thu, 28 Jul 2005, Tracy Wang wrote: Hi Doug, thanks for answering so quickly. Actually i ran recon-all -subjid 000112_vc1552_test -autorecon1 which i think would include the initial talairaching stage correct? the talairach tranform seemed problematic both when viewed in tkmedit and the final objective funtion value in the mritotal.log (ours was 0.15870732). does this mean that the talairaching failed at some place? How can i tell because mritotal seems to run to completion? Thanks again, Tracy On Jul 28, 2005, at 4:52 PM, Doug Greve wrote: Tracy Wang wrote:Hi all, I'm trying to run tkregister2 and I have the following questions: 1. It seems to look for a talairach directory in the main subject directory. Does this mean I have to make/modify this directory each time i want the tkregister2 to manually correct the talairach transform? ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ talairach.tar.gz. It means that you did not run the initial talairaching stage (eg, recon-all -talairach). If you don't want to run this first, you can run tkregister2 with --regheader (but be careful to run subsequent calls to recon-all with -notalairach to keep from overwriting). Also note that you can get help with tkregister2 --help doug tkregister2 --mgz --s 000112_vc1552_test --fstal --surf orig surfname set to orig INFO: no target volume specified, assuming FreeSurfer orig volume. target volume orig movable volume (null) reg file (null) LoadVol 1 $Id: tkregister2.c,v 1.35 2005/07/12 19:50:28 greve Exp $ Diagnostic Level -1 talairach.xfm - 0.874 -0.173 -0.004 -3.334; 0.188 0.957 -0.284 -10.880; 0.046 0.217 0.760 31.928; 0.000 0.000 0.000 1.000; WARNING: can't find the talairach xform '/Users/Shared/freesurfer/subjects/00112_vc1552_test/mri/transforms/ talairach.xfm' WARNING: transform is not loaded into mri corRead(): can't open file /Users/Shared/freesurfer/subjects/talairach/mri/orig/COR-.info ERROR: could not read /Users/Shared/freesurfer/subjects/talairach/mri/orig as 0 Thanks for any assistance! Tracy Wang Research Assistant Cognitive Neuroscience Laboratory Washington University Department of Psychology, Campus Box 1125 One Brookings Drive St. Louis MO 63130-4899 314-935-5019 [EMAIL PROTECTED] ___ 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 [EMAIL PROTECTED] Phone Number: 617-724-2358 Fax: 617-726-7422 Tracy Wang Research Assistant Cognitive Neuroscience Laboratory Washington University Department of Psychology, Campus Box 1125 One Brookings Drive St. Louis MO 63130-4899 314-935-5019 [EMAIL PROTECTED] ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] stage1
Yeah, usually it took longer when there was something that went wrong. I suspect that the skull-stripping step didn't go well. Please check the "brain" volume in tkmedit. Xiao Han On Mon, 18 Jul 2005, Timothy O'Keefe wrote: I ran recon-all stage1 and it took a very long time. Well, at least longer than people document. I didn't time it, but it must have taken close to 10 hours. Also I received this error during mri_fill(): ... using +/- offset search region where offset is 39. using +/- offset search region where offset is 42. using +/- offset search region where offset is 45. using +/- offset search region where offset is 48. mri_fill: could not find pons No such file or directory INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst talairach cc position changed to (0.90, -7.51, 23.75) INFO: Modifying dst c_(r,a,s), using the transform dst INFO: Modifying dst c_(r,a,s), using the transform dst find_cutting_plane:seed point not in structure! Searching neighborhood... find_cutting_plane:seed point not in structure! Searching neighborhood... ERROR: mri_fill exited with non-zero status recon-all exited with errors at Sat Jul 16 01:13:52 EDT 2005 Are there any suggestions? I am using a single T1 (1x1x1) image acquired from a Siemens 1.5T machine. I realize I should use multiple images for averaging, however I am just trying to get a feel for the software. -T ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] Failed segmentation
Hi Akram, In that case, you will need to manually edit the WM volume to separate cerebellum from WM. You can load both WM and filled volume to tkmedit, and go through the slices to see where in "filled" volume cerebellum is connected to WM, and then edit the WM to cut off the connections. Then, save the edited WM to replace the old one. After the editing of WM, rerun mri_fill (or -stage2 of recon-all). Hope that helps. Xiao Han On Tue, 12 Jul 2005, Bakkour, Akram wrote: Dear all, I am working on a recent MPRAGE image obtained at MGH. First, we had trouble converting the files, but Dr. Ress converted it to analyze manually, I then used that analyze file in freesurfer. The image has a lot of inhomogeneity and the intensity is variable from the back and front of the brain and also between hemispheres. wmfill includes the cerebellum as RH even if I set the coordinates for the cutting planes in expert preferences. Any suggestions as to what I should try next? Thanks for the help! Regards, Akram. ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] recon-all-nmr -disable-autoseg
I think you need to put -disable-autoseg in front of -stage1. -Xiao On Tue, 7 Jun 2005, Elizabeth Fenstermacher wrote: Hello All, I have a number of jobs that I will be running through recon-all-nmr -stage1 in the next few days. I tried using the -disable-autoseg flag, but it seems not to work as the jobs I submitted this morning have now run for 50 minutes. The command line as I typed it was pbsubmit -c "recon-all-nmr -stage1 -subjid $s -nuintensitycor -usenuintensitycor -disable-autoseg" perhaps I've misplaced the flag? If anyone has any insight as to why it seems to be persitting in running the autoseg I would appreciate it. Thanks! Eliz ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] wm fill won't work with partial brain
Hi Akram, One possible way is to manually set the Seed points for corpus callosum etc using the -C, -P, -lh, and -rh options of mri_fill. Note that these coordinates must be in the Talairach space. -Xiao On Mon, 23 May 2005, Bakkour, Akram wrote: Hello, I have a high resolution image covering the back third of the brain obtained from surface coils and wm fill fails, it tells me that it doesn't find the corpus callosum. This makes sense because it is not a full brain. But I was wondering if there was a way around this and have the program continue to make progress rather than fail and stop. Thanks for the help! Akram. ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] average curvature on individual subjects
It is possible. You can try mri_surf2surf. -Xiao On Wed, 18 May 2005, Frida Polli wrote: Hi I just learned that it is possible to make an average curvature file from a bunch of subjects, but specify the output subject as an individual brain. given that, is it possible to put average7's curvature files onto an individual subject? thanks, frida ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
[Freesurfer] Re: Parcellation IV
So, it seems you do need to update your binaries. The correspondence between label and color is specified by the text file tkmeditParcColorsCMA in the FREESURFER_HOME directory, I believe. But after you load the annotation into tksurfer and move the cursor over each region, the anatomical name should be displayed by the program. -Xiao On Tue, 22 Mar 2005, Wan Park wrote: Thanks Xiao Han! I worked finally! Cheers (see below) Now I have this question: I see there is a lot of regions. How do each region correspond to specific color in the visualization? Thanks a lot! Wan total white matter volume = 533145 mm^3 table columns are: number of vertices total surface area (mm^2) total gray matter volume (mm^3) average cortical thickness +- standard error (mm) integrated rectified mean curvature integrated rectified Gaussian curvature folding index intrinsic curvature index structure name 2955 1803 6322 3.453 +- 0.875 0.204 0.080 71.570 9.042 G_parietal_inferior-Angular_part 3572 2241 6032 2.691 +- 0.894 0.192 0.060 59.096 9.520 S_intraparietal-and_Parietal_transverse 2379 1388 3665 2.599 +- 0.657 0.190 0.060 58.310 6.513 G_parietal_superior 1269713 1901 2.621 +- 1.066 0.208 0.087 33.729 5.435 G_occipital_middle 824464 1105 2.449 +- 1.013 0.159 0.057 14.657 1.930 S_occipital_superior_and_transversalis 5746 3588 9688 2.721 +- 0.785 0.153 0.043 60.049 8.714 S_temporal_superior 3950 2272 5219 2.285 +- 0.617 0.183 0.060 74.201 8.967 G_precentral 3509 1947 5172 2.636 +- 0.791 0.193 0.078 291.730 9.736 G_postcentral 4558 2846 6608 2.321 +- 0.553 0.170 0.043 54.653 7.641 S_central 3180 1931 5328 2.757 +- 0.611 0.147 0.037 30.201 4.109 S_postcentral 2393 1395 2388 1.717 +- 0.559 0.225 0.074 58.455 7.835 G_paracentral 465248684 2.720 +- 0.584 0.138 0.0405.397 1.079 S_intermedius_primus-Jensen 2626 1508 3646 2.381 +- 0.738 0.158 0.047 37.750 4.537 G_precuneus 2486 1441 3440 2.371 +- 0.824 0.214 0.088 68.986 9.620 Pole_occipital 1150831 1277 1.559 +- 0.531 0.201 0.059 13.629 3.664 G_occipital_superior 1045613 1346 2.158 +- 0.733 0.191 0.069 20.900 2.763 G_occipital_inferior 2403 1278 3600 2.793 +- 0.657 0.182 0.067 52.261 6.180 G_temporal_middle 1531921 2238 2.456 +- 0.524 0.197 0.055 26.134 2.962 S_precentral-Superior-part 1747926 2300 2.469 +- 0.746 0.165 0.065 26.462 4.279 G_parietal_inferior-Supramarginal_part 788441 1023 2.287 +- 0.956 0.205 0.097 16.175 3.575 S_occipital_middle_and_Lunatus 1694924 1989 2.253 +- 0.946 0.204 0.087 47.149 7.090 G_cuneus 301118243 2.046 +- 0.736 0.140 0.0594.324 0.713 S_occipital_inferior 770453 1140 2.513 +- 0.520 0.162 0.0418.611 1.256 S_occipital_anterior 8169 4643 9060 1.971 +- 0.606 0.185 0.061 142.550 19.432 G_frontal_superior 806483904 1.868 +- 0.652 0.170 0.054 11.111 1.823 S_cingulate-Marginalis_part 761454869 1.961 +- 0.665 0.145 0.0377.589 0.992 S_subparietal 2634 1496 3645 2.456 +- 0.629 0.205 0.076 62.685 6.930 G_temporal_inferior 385173500 2.845 +- 0.535 0.133 0.0414.932 0.410 G_temp_sup-Planum_tempolale 946577 1356 2.365 +- 0.443 0.172 0.0409.745 1.653 S_temporal_inferior 2556 1766 2644 1.539 +- 0.548 0.174 0.045 35.471 6.396 S_parieto_occipital 1608945 2243 2.355 +- 0.613 0.154 0.035 12.523 2.516 Lat_Fissure-post_sgt 2976 1816 4799 2.656 +- 0.539 0.209 0.061 67.117 7.108 G_occipit-temp_lat-Or_fusiform 91 42 56 1.341 +- 0.194 0.141 0.0260.851 0.105 S_paracentral 2519 1720 3436 2.049 +- 1.045 0.152 0.040 26.893 5.123 S_calcarine 365212345 1.644 +- 0.419 0.128 0.0152.299 0.248 S_collateral_transverse_post 1247714 1541 2.183 +- 0.494 0.165 0.036 12.002 2.265 S_occipito-temporal_lateral 3489 2058 3863 1.885 +- 0.594 0.175 0.049 48.747 6.589 S_cingulate-Main_part 1792 1174 2649 2.327 +- 1.109 0.179 0.041 38.026 2.807 G_occipit-temp_med-Lingual_part 1340746 1715 2.317 +- 0.546 0.177 0.061 22.078 3.527 G_temp_sup-Lateral_aspect 3040 1843 3431 1.874 +- 0.487 0.147 0.031 28.439 3.845 S_frontal_superior 1638893 2173 2.380 +- 0.695 0.159 0.045 24.985 3.034 G_cingulate-Main_part 637336867 2.556 +- 0.682 0.171 0.064 14.747 1.730 G_subcentral 214102226 2.234 +- 0.679 0.080 0.0170.636 0.115