The FreeSurferColorLUT.txt file has over 1,200 regions listed in it, but I assume that not all of those could be output in those default recon-all aparc+aseg* files. For example, I don't think I have ever seen a region: "3001 wm-lh-bankssts 230 155 215 0" output.
So, I assume that there exists some set S of size N<1200 regions *could* be output in standard recon-all output aparc+aseg and aparc+aseg2009 files. I am interested in having a list of regions in S.
You provided a list of region numbers for WM+CSF that can be expected in standard recon-all output aparc+aseg and aparc+aseg2009. So, those regions are part of what I am calling S, and presumably the rest of the regions in S are all GM ROIs, since I can subtract the list of WM+CSF regions from the aparc+aseg* outputs and have a GM mask (-> what "mri_binarize --gm ..." does).
I was trying to guess what portions of the FreeSurferColorLUT.txt table might form S, by looking at the groupings of regions presented in the file, as well as the list of WM+CSF regions you had provided and some recon-all output files that I had.
So, I was wondering if S is made up of the following regions: numbers [0-255] and [1000-2035] for aparc+aseg (i.e., lines 5-231 and 430-502 in the LUT), and numbers [0-255], [1000-2035] and [11100-12175] for aparc+aseg2009 (i.e., lines 5-231, 430-502 and 1089-1241 in the LUT) ?
Thanks, pt
On Wed, Nov 9, 2016 at 9:41 AM, Douglas Greve greve@nmr.mgh.harvard.edu wrote:
Sorry, I'm not sure what your question is here?
On 11/8/16 5:25 PM, P Taylor wrote:
OK, thanks for that list of complementary ROIs.
I've always wondered what is the maximal set of regions that can be returned by the standard recon-all with no other options for aparc+aseg and aparc+aseg2009? Looking at the FreeSurferColorLUT.txt, I guess there is a batch from lines 1-231 (spanning integer range 0-255), then some groupings of regions get skipped, picking up again at ~line 415 with the comment about mri_aparc2aseg (integer range 1000-2035, the max of which seems to be the largest int in the aparc+aseg), and then is there a jump again to line 1079, starting with the comment about the --a2009s flag (integer range 11100-12175, the last of which seems to be the largest int in the aparc+aseg2009)?
Is that a correct set of regions that are returned? And out of that combined set, we can deselect any of the "MatchValues" of WM/CSF/background/etc. to be left with GM?
Thanks, pt
On Tue, Nov 8, 2016 at 1:32 PM, Douglas Greve greve@nmr.mgh.harvard.edu wrote:
Actually, what it does is to binarize all the labels that it thinks are not GM, and then invert the mask to give GM. for what it is worth, there are the non-GM numbers MatchValues[nMatch++] = 2; MatchValues[nMatch++] = 41; MatchValues[nMatch++] = 77; MatchValues[nMatch++] = 251; MatchValues[nMatch++] = 252; MatchValues[nMatch++] = 253; MatchValues[nMatch++] = 254; MatchValues[nMatch++] = 255; MatchValues[nMatch++] = 7; // Cerebellar WM MatchValues[nMatch++] = 46; // Cerebellar WM MatchValues[nMatch++] = 4; // Left-Lateral-Ventricle MatchValues[nMatch++] = 5; // Left-Inf-Lat-Vent MatchValues[nMatch++] = 14; // 3rd-Ventricle MatchValues[nMatch++] = 43; // Right-Lateral-Ventricle MatchValues[nMatch++] = 44; // Right-Inf-Lat-Vent MatchValues[nMatch++] = 15; // 4th-Ventricle MatchValues[nMatch++] = 72; // 5th-Ventricle MatchValues[nMatch++] = 31; // Left-choroid-plexus MatchValues[nMatch++] = 63; // Right-choroid-plexus MatchValues[nMatch++] = 0; // Background MatchValues[nMatch++] = 24; // CSF
On 11/8/16 1:06 PM, P Taylor wrote:
OK. But going back to my question that kicked this thread off, actually what I was looking for was just such a rule.
I could make a fairly solid list of GM from using this new "mri_binarize --gm ...." on several recon-all outputs. Would it be possible to have the rule (presumably a list of numbers corresponding to GM ROIs?) so I can have as secure a set of GM labels as possible?
Thanks, pt
On Tue, Nov 8, 2016 at 12:29 PM, Douglas Greve <greve@nmr.mgh.harvard.edu
wrote:
Yes. It just uses a rule on the numbers that it expects to be GM. At some point, I'll add both tissue type and laterality info to the ROIs, but not for version 6
On 11/8/16 12:03 PM, P Taylor wrote:
And a final question, which might be a silly one-- This GM map applies to both the aparc.a2009s+aseg.mgz and aparc+aseg.mgz files, right?
Running "mri_binarize --gm ..." on each of those files separately gave the same mask-- I believe this should be a general property?
Thanks, pt
On Tue, Nov 8, 2016 at 11:57 AM, Douglas Greve < greve@nmr.mgh.harvard.edu> wrote:
yes
On 11/8/16 11:51 AM, P Taylor wrote:
Ah, cool, that looks like it does the trick-- thanks much.
Will that version of mri_binarize be in the next release of FS?
Thanks, pt
On Tue, Nov 8, 2016 at 11:35 AM, Douglas Greve < greve@nmr.mgh.harvard.edu> wrote:
Try this one https://gate.nmr.mgh.harvard.e du/safelinks/greve/mri_binarize
On 11/8/16 11:26 AM, P Taylor wrote:
Thanks, I'd like to try that. However, my mri_binarize doesn't show that option, and when I try running it, that option isn't recognized:
$ mri_binarize --gm --o map_gm.nii --i aparc+aseg.mgz ERROR: Option --gm unknown
Looking at the mri_binarize help, I see '--wm' exists, and that does, indeed, produce a WM-looking volume. I also see an option "--wm+vcsf"; would subtracting that resulting map leave only GM?
My version of FS is:
You are running this version of FreeSurfer:
freesurfer-Linux-centos6_x86_64-stable-pub-v5.3.0
In the online description of mri_binarize: https://surfer.nmr.mgh.harvard.edu/fswiki/mri_binarize I see a "--subcort-gm" option that doesn't appear to work for me, but I guess that isn't full WM.
So, am I somehow behind the times? From the download page: https://surfer.nmr.mgh.harvard.edu/fswiki/DownloadAndInstall it looks like my v5.3 is the most currently available?
... and, in case it's useful, the explicit version of my mri_binarize (bottom line of the help) is: $Id: mri_binarize.c,v 1.26.2.1 2011/04/08 15:40:50 greve Exp $
Thanks, pt
On Tue, Nov 8, 2016 at 11:08 AM, Douglas Greve < greve@nmr.mgh.harvard.edu> wrote:
We don't currently have that information (ie, tissue type) stored with the ROI. One thing you can do is to run mri_binarize with the --gm option on aparc+aseg.mgz This uses some rules to determine what is GM and creates a mask of GM. You can then feed this into mri_segstats with --mask, and it should only return ROIs that are in the GM mask.
On 11/7/16 8:10 PM, P Taylor wrote:
Thanks, that's useful for getting the ROI volumes and a list of all that's been found for a particular data set.
I would still like to have a way to identify a listed region as gray matter or not so that I can parse the file for only GM regions that have been returned, particularly as different numbers of ROIs are typically returned in an aparc+aseg file.
I see in the aseg.stats file that there are volumetric summations for GM quantities, such as: # Measure lhCortex, lhCortexVol, Left hemisphere cortical gray matter volume, 238183.759207, mm^3 # Measure rhCortex, rhCortexVol, Right hemisphere cortical gray matter volume, 240824.502797, mm^3 # Measure Cortex, CortexVol, Total cortical gray matter volume, 479008.262004, mm^3 .... # Measure SubCortGray, SubCortGrayVol, Subcortical gray matter volume, 57815.000000, mm^3 # Measure TotalGray, TotalGrayVol, Total gray matter volume, 643956.262004, mm^3
Are these calculated by summing individual ROI volumes, so that a list of all possible GM volumes does already exist somewhere?
Thanks, pt
On Mon, Nov 7, 2016 at 3:39 PM, Douglas N Greve < greve@nmr.mgh.harvard.edu> wrote:
> I don't know that we have a complete list in one place of just the > aparc+aseg ROIs, but you can get one by running > > mri_segstats --seg aparc+aseg.mgz --ctab-default --sum sum.dat > > then look in the sum.dat file > > > On 11/03/2016 04:32 PM, P Taylor wrote: > > Is there an available list of GM regions within the list of > regions in > > the FreeSurferColorLUT.txt? > > > > (In particular, I am interested in a list of GM regions output in a > > default run of recon-all, which seems to have a maximum > regionnumber > > of 2035. But a full list is great, too-- I'm not sure which might > be > > already available or easier to generate.) > > > > I had thought of using the segmentation volumes to help me > > differentiate ROIs. For example, I could get rid of regions that > > overlap with the WM volume, but I think that the segmentation might > > come from a slightly different bit of information than the sum of > > segmented regions. As shown in the attached slice image, if I > > underlay the WM segmentation volume (color = white), and overlay a > > "recon-all" parcellation+segmentation map (color = red for GM on > > right, yellow for GM on left, and green for non-GM) to look for > > overlap, I see that some of the WM seg volume overlaps with GM ROIs > > (overlaps are isolated pink-ish voxels and light yellow voxels; one > > highlighted in cross hairs). I guess this difference occurs > because > > the segmentation volume comes from the surfaces themselves? > > > > Anyways, this makes me think that I can't automate finding the GM, > and > > I would rather go from LUT numbers, and hence the above question. > > > > Thanks, > > pt > > > > > > > > _______________________________________________ > > 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: https://gate.nmr.mgh.harvard.edu/filedrop2 > www.nmr.mgh.harvard.edu/facility/filedrop/index.html > Outgoing: ftp://surfer.nmr.mgh.harvard.e > du/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 listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.e du/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 listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.e du/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 listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.e du/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 listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.e du/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 listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.e du/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 listFreesurfer@nmr.mgh.harvard.eduhttps://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
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.