Re: [Freesurfer] Merging aparc+aseg.mgz and brainstemSsLabels.v10.FSvoxelSpace.mgz
External Email - Use Caution Hi Eugenio, Thank you for your reply. The solution to blend the two segmentation will work for me. If you could share the Matlab code that would be great. Also one other related question i have is related to edits and brainstem-structures. 1. The first time i run recon-all, i use: "recon-all -all -s bert -brainstem-structures. 2. Then suppose, I add control points or edit brain-mask; i re-run recon-all with: 3. recon-all -s bert -autorecon2-cp -autorecon3 1. when i do this i think i should also add the brainstem-structures flag again. thus my command will be: 4. recon-all -s bert -autorecon2-cp -autorecon3 -brainstem-structures. Does step-4 look correct to you ? Thanks Mehul On Thu, Jun 21, 2018 at 2:44 AM, Iglesias Gonzalez, Eugenio < e.igles...@ucl.ac.uk> wrote: > Dear Mehul, > > Yes, the labels from the Brainstem module overlap with the definition of > ventral-dc in aseg.mgz, since they are based on atlas built with > delinations made with different criteria. > > A trick you can use to blend the two segmentations into one is: > > 1. Create a binary mask for the whole brainstem using > brainstemSsLabels.v10.FSvoxelSpace.mgz. > > 2. In aseg.mgz (or aparc+aseg.mgz), replace the label of every voxel in > the mask by that of the nearest voxel outside the mask. > > 3. Paint the brainstem from brainstemSsLabels.v10.FSvoxelSpace.mgz over > the mask in the output of step 2. > > Does this make any sense? I can provide you with some Matlab code if you > with. > > Cheers, > > /E > > > > -- > > Juan Eugenio Iglesias > > ERC Senior Research Fellow > > Centre for Medical Image Computing (CMIC) > > University College London > > http://www.jeiglesias.com > > http://cmictig.cs.ucl.ac.uk/ > > > > > > *From: *Mehul Sampat > *Date: *Thursday, 21 June 2018 at 04:10 > *To: *Freesurfer Mailing List , "Iglesias > Gonzalez, Eugenio" > *Subject: *Merging aparc+aseg.mgz and brainstemSsLabels.v10. > FSvoxelSpace.mgz > > > > Hi Eugenio and Bruce, > > > > We used freesurfer-6.0.0 to segment the brain-stem into different > different structures. > > > > I see we get two mgz files after this process: > > aparc+aseg.mgz and brainstemSsLabels.v10.FSvoxelSpace.mgz > > > > Do you have any recommendations to combine these two masks into a single > mask ? > > > > I see pons in brainstemSsLabels.v10.FSvoxelSpace.mgz overlaps with the > brain-stem in aparc+aseg.mgz but the midbrain in > brainstemSsLabels.v10.FSvoxelSpace.mgz > overlaps with brain-stem and the ventral-dc in aparc+aseg.mgz. > > > > Any recommendations are appreciated. > > Thanks > > Mehul > > > > > > > > ___ 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] Merging aparc+aseg.mgz and brainstemSsLabels.v10.FSvoxelSpace.mgz
External Email - Use Caution Hi Eugenio and Bruce, We used freesurfer-6.0.0 to segment the brain-stem into different different structures. I see we get two mgz files after this process: aparc+aseg.mgz and brainstemSsLabels.v10.FSvoxelSpace.mgz Do you have any recommendations to combine these two masks into a single mask ? I see pons in brainstemSsLabels.v10.FSvoxelSpace.mgz overlaps with the brain-stem in aparc+aseg.mgz but the midbrain in brainstemSsLabels.v10.FSvoxelSpace.mgz overlaps with brain-stem and the ventral-dc in aparc+aseg.mgz. Any recommendations are appreciated. Thanks Mehul ___ 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] using -bigventricles flag in FS 5.3
Hi Folks, I have a question about the usage of the -bigventricles flag in FS 5.3. 1. Suppose i run recon-all with -bigventricles option. 2. Then i add control points and run recon-all with the -autorecon2-cp as described in the recon-all help. USAGE: recon-all 3. The recon-all help mentions "-autorecon2-cp : process stages 12-23 (uses -f w/ mri_normalize, -keep w/ mri_seg" My question is: when i run recon-all the second time with the -autorecon2-cp flag, do i need to include the -bigventricles option again ? In general, if the first round of processing used -bigventricles, should it be always used if recon-all is re-run after making edits ? Thanks Mehul ps: can you believe the Patriots won SB51 !! :-) ___ 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] running dt_recon on a philips Achieva DTI dataset;
Hi Folks, I am trying to run dt_recon on a Philips Achieva DTI dataset. I could not find an example of a bvecs for this dataset so i made one in this way: I have 850 dicom slices. mri_convert correctly splits it into 17 nifti files (50 slices each). I took 17 dicom slices with the same ImagePositionPatient and then found the DiffusionGradientOrientation for these 17 slices (the values are below). I just wanted to check that: 1. These are the 17 vectors I should use in the bvecs file for dt_recon ? 2. This is the order i should keep these vectors in ? 3. I thought there would be 17 unique Diffusion GradientOrientation values; but I only find 16 in the slices (0 0 0) is repeated twice. (One of the slices with (0,0,0) has a bvalue of 0 and while the other has a b-value of 2500) Mehul 0 0 0 -0.5000 -0.5000 -0.7071 -0.5000 -0.50000.7071 0.7071 -0.70710. -0.1104 -0.7070 -0.6985 0.2893 -0.6996 -0.6534 0.6275 -0.3305 -0.7050 0.6574 -0.2734 -0.7022 -0.6725 -0.5421 -0.5038 0.7038 -0.49110.5133 -0.6930 -0.25310.6751 -0.7066 -0.70710.0277 -0.2821 -0.70700.6485 0.2886 -0.70160.6514 0.7058 -0.70630.0537 0.7014 -0.20500.6827 0 0 0 ___ 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] Which Freesurfer regions in Default mode network ?
Hi Folks, I wanted to ask which Freesurfer regions from aparc+aseg.mgz are included in Default mode network ? Is there a relevant publication on this topic ? Thanks Mehul ___ 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.
Re: [Freesurfer] missing cortical parcellation labels in aparc+aseg
Hi Bruce, rh.aparc.stats looks good and also rh.aparc.a2009s.stats looks correct. However, both aparc+aseg.mgz and aparc.a2009s+aseg.mgz have the issue with missing parcellation labels in right hemisphere. Also I suspect the issues starts occurring in rh.ribbon.mgz and ribbon.mgz (missing labels). (lh.ribbon.mgz is fine) I tried to recreate the rh.ribbon.mgz and ribbon.mgz files using the command below and I am able to reproduce the issue. mris_volmask http://freesurfer.net/fswiki/mris_volmask --label_left_white 2 --label_left_ribbon 3 --label_right_white 41 --label_right_ribbon 42 --save_ribbon --save_distance subjid I can upload the data if you would like to take a look. Thanks Mehul On Fri, Aug 15, 2014 at 1:54 AM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: How does the rh.aparc look? On Aug 14, 2014, at 8:13 PM, Mehul Sampat mpsam...@gmail.com wrote: Hi Folks, We have a strange error for one case. For one case there are missing cortical parcellation labels in aparc+aseg.mgz file in the right hemisphere (see screenshots) I looked at the recon-all.log file and no errors are reported. Also I recreated aparc+aseg.mgz file using the following command but the issue persists: mri_aparc2aseg --s subjid --volmask I looked at the rh.white and rh.pial and those look good. I cannot figure out what would cause such an issue ? I was wondering if anyone has seen something similar ? Thanks Mehul missing-parcellation-10591037.png s3061-s10591037-img3.png s3061-s10591037-img2.png 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.
Re: [Freesurfer] missing cortical parcellation labels in aparc+aseg
Thank Bruce. I have uploaded the case to transfer/incoming in surfer.nmr.mgh.harvard.edu My file is missingRightParcLabels.tar.gz Mehul On Fri, Aug 15, 2014 at 3:17 PM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: Sure, upload it and one of us will take a look Bruce On Aug 15, 2014, at 10:28 me PM, Mehul Sampat kmpsam...@gmail.com wrote: Hi Bruce, rh.aparc.stats looks good and also rh.aparc.a2009s.stats looks correct. However, both aparc+aseg.mgz and aparc.a2009s+aseg.mgz have the issue with missing parcellation labels in right hemisphere. Also I suspect the issues starts occurring in rh.ribbon.mgz and ribbon.mgz (missing labels). (lh.ribbon.mgz is fine) I tried to recreate the rh.ribbon.mgz and ribbon.mgz files using the command below and I am able to reproduce the issue. mris_volmask http://freesurfer.net/fswiki/mris_volmask --label_left_white 2 --label_left_ribbon 3 --label_right_white 41 --label_right_ribbon 42 --save_ribbon --save_distance subjid I can upload the data if you would like to take a look. Thanks Mehul On Fri, Aug 15, 2014 at 1:54 AM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: How does the rh.aparc look? On Aug 14, 2014, at 8:13 PM, Mehul Sampat mpsam...@gmail.com wrote: Hi Folks, We have a strange error for one case. For one case there are missing cortical parcellation labels in aparc+aseg.mgz file in the right hemisphere (see screenshots) I looked at the recon-all.log file and no errors are reported. Also I recreated aparc+aseg.mgz file using the following command but the issue persists: mri_aparc2aseg --s subjid --volmask I looked at the rh.white and rh.pial and those look good. I cannot figure out what would cause such an issue ? I was wondering if anyone has seen something similar ? Thanks Mehul missing-parcellation-10591037.png s3061-s10591037-img3.png s3061-s10591037-img2.png 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.
[Freesurfer] papers on reliability of volume of indivdual cortical parcellation regions
Hi Folks, I was able to find a number of papers looking at the reliability of cortical thickness of individual regions (ex: https://surfer.nmr.mgh.harvard.edu/pub/articles/reliability_wonderlick.pdf) I was wondering if there are any papers that have looked at the reliability of the volumes of the individual cortical parcellation regions ? Thanks Mehul ___ 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.
Re: [Freesurfer] Warping to MNI space
Hi Doug, I have a followup question on this topic. I understand the two command you mentioned in this thread to warp a segmentation to the MNI152 space. In the subjects dir, I see there is an average subject cvs_avg35_inMNI152 which is already in MNI space (and the volume size is 256^3). I want to just reslice the aparc+aseg.mgz to size of the MNI template (91*109*91) Since this subject is already in MNI152 space, I can just do the following right? mri_label2vol --subject cvs_avg35_inMNI152 --aparc+aseg --temp spm-T1.nii --o aparc+aseg2.nii --regheader aparc+aseg.mgz (where spm-T1.nii is the T1 template from spm) Or would you recommend a different approach ? Thanks Mehul ps: Earlier I used mri_convert as follows: mri_convert aparc+aseg.mgz -rt nearest -rl spm-T1.nii aparc+aseg1.mgz But then I recalled your advice (from earlier) to use mri_label2vol convert the segmentation labels and followed it! On Mon, Nov 25, 2013 at 3:46 PM, Douglas N Greve gr...@nmr.mgh.harvard.eduwrote: Note: you can run each command with --help to get info something like mni152reg --s subject to create the registration file (it will go into $SUBJECTS_DIR/$subject/mri/transforms/reg.mni152.2mm.dat then something like mri_label2vol --seg $SUBJECTS_DIR/$subject/mri/aseg.mgz --reg $SUBJECTS_DIR/$subject/mri/transforms/reg.mni152.2mm.dat --o aseg.$subject.mni152.nii doug On 11/25/2013 11:09 AM, ebell...@uwm.edu wrote: Hi Douglas, Thanks for the reply. I am very new to Freesurfer and am more used to scripting with AFNI. Would you mind giving me a bit of sample code to get used to what this may look like? Can I script this in a .tcsh file? Also, once I convert the aseg to MNI space, can I simply average or add the individual masks (which could be done in afni) or is there a better way to do this in Freesurfer? Also, is this a linear or non-linear warp? Thanks, Emily - Original Message - From: Douglas N Greve gr...@nmr.mgh.harvard.edu To: freesurfer@nmr.mgh.harvard.edu Sent: Thursday, November 21, 2013 12:57:07 PM Subject: Re: [Freesurfer] Warping to MNI space Do you mean the aseg.mgz for an individual? If so, you can run mni152reg to register that subject to mni152space. Next you can run mri_label2vol to map the aseg to the 152 space. doug On 11/17/2013 07:26 PM, ebell...@uwm.edu wrote: Hello, I plan on using the aseg.nii to create a grey matter mask. I want to warp these grey matter masks to MNI space. I have been trying to use @auto_tlrc in AFNI to do this warping but am having some trouble. The reason I need this in MNI space is because I plan to make a group grey matter mask that I use to restrict my funtional fmri data analysis, which is in MNI space (MNI_avg_152T1). How might I go about doing the warping and I wanted to double check the order of doing things? Should I create the grey matter masks first then warp to MNI? Or do I warp the aseg files to MNI and then create the grey matter masks? Thanks so much! Emily ___ 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 gr...@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.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] regarding adding and deleting voxels in wm.mgz and brainmask.mgz
Hi Folks, I see on the tutorial websites that when voxels are added to wm.mgz or brainmask.mgz, the new voxels are given a value of 255; while when voxels are deleted, these locations are set to 1. https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/WhiteMatterEdits Can the deleted voxels can be set to 0 or they have to be set to 1; Or is 0 used to differentiate the background voxels from brain voxels ? Thanks Mehul ___ 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] question about format of control points
Hi Folks, Once we make control points in tkmedit, they are stored in the file control.dat Are these control points stored in RAS co-ordinate system ? If not what is the co-ordinate system used to store the control points Thanks Mehul ___ 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.
Re: [Freesurfer] question about format of control points
Thanks Bruce; I noticed that in the control.dat file the second last line is useRealRAS 0 just curious to know if info is used by recon-all or is only used by the viewer ? Thanks Mehul On Mon, Jul 1, 2013 at 11:34 AM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: yes, they are stored in ras On Mon, 1 Jul 2013, Mehul Sampat wrote: Hi Folks, Once we make control points in tkmedit, they are stored in the file control.dat Are these control points stored in RAS co-ordinate system ? If not what is the co-ordinate system used to store the control points Thanks Mehul 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/**compliancelinehttp://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.
Re: [Freesurfer] question about format of control points
Thanks Bruce; Okay i assume, useRealRAS should be 1 for scanner RAS; Two questions: 1. If my control points are in scanner RAS, then can I use them directly as long as I set useRealRAS to 1 in the control.dat file ? 2. Can useRealRAS variable be only 0/1 or do you have a different value for surface RAS? Thanks Mehul On Mon, Jul 1, 2013 at 12:13 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Mehul they can either be in real (meaning scanner) RAS or in surface RAS cheers Bruce On Mon, 1 Jul 2013, Mehul Sampat wrote: Thanks Bruce; I noticed that in the control.dat file the second last line is useRealRAS 0 just curious to know if info is used by recon-all or is only used by the viewer ? Thanks Mehul On Mon, Jul 1, 2013 at 11:34 AM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: yes, they are stored in ras On Mon, 1 Jul 2013, Mehul Sampat wrote: Hi Folks, Once we make control points in tkmedit, they are stored in the file control.dat Are these control points stored in RAS co-ordinate system ? If not what is the co-ordinate system used to store the control points Thanks Mehul 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/**compliancelinehttp://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.
Re: [Freesurfer] question about format of control points
Thanks Bruce. Okay will convert them from scanner RAS to surface RAS; last question: I looked at the control points made from tkmedit; it seems like you round of the co-ordinates to integers and then save them as floats; Does the co-ordinate information saved in control.dat need to be rounded or is it okay to save floats in control.dat; ? I also noticed the following difference between tkmedit and freeview; if I make and save control points in tkmedit it says useRealRAS 0 while if i make and save control points in freeview it says useRealRAS 1 (here the co-ordinate info is not rounded off). Not sure if this could affect the workflow but just wanted to let you know; Thanks Mehul ps: In case, anyone needs to convert from scanner RAS to surface RAS; this is how I am doing it in Matlab: 1. First read the vox2ras0 from MRIread and compute it's inverse. 2. Then do a matrix multiplication to get surface RAS. that is: surface RAS = inv(vox2ras0) * scanner RAS; Mehul On Mon, Jul 1, 2013 at 12:51 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: 1. Yes, I think so, although I wouldn't swear that everything is smart enough to read this flag. You might be safer converting them to surface ras and using them that way 2. Yes. On Mon, 1 Jul 2013, Mehul Sampat wrote: Thanks Bruce; Okay i assume, useRealRAS should be 1 for scanner RAS; Two questions: 1. If my control points are in scanner RAS, then can I use them directly as long as I set useRealRAS to 1 in the control.dat file ? 2. Can useRealRAS variable be only 0/1 or do you have a different value for surface RAS? Thanks Mehul On Mon, Jul 1, 2013 at 12:13 PM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: Hi Mehul they can either be in real (meaning scanner) RAS or in surface RAS cheers Bruce On Mon, 1 Jul 2013, Mehul Sampat wrote: Thanks Bruce; I noticed that in the control.dat file the second last line is useRealRAS 0 just curious to know if info is used by recon-all or is only used by the viewer ? Thanks Mehul On Mon, Jul 1, 2013 at 11:34 AM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: yes, they are stored in ras On Mon, 1 Jul 2013, Mehul Sampat wrote: Hi Folks, Once we make control points in tkmedit, they are stored in the file control.dat Are these control points stored in RAS co-ordinate system ? If not what is the co-ordinate system used to store the control points Thanks Mehul 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/**compliancelinehttp://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] creating an entorhinal cortex volume from ?h.entorhinal_exvivo.label
Hi Folks, I am trying to create an entorhinal cortex volume from ?h.entorhinal_exvivo.label using mri_label2vol I cd to the subject dir ($SUBJECTS_DIR/bert) and the command I use is: *mri_label2vol --label label/lh.entorhinal_exvivo.label --temp mri/nu.mgz --identity --o mri/lh-new-erc.mgz * I took the sum of the voxels in this new mask i created (lh-new-erc.mgz) and it is 215; I looked at the volume provided in lh.entorhinal_exvivo.stats and it is 925; (from the log file I believe this is created using the command: mris_anatomical_stats -mgz -f ../stats/lh.entorhinal_exvivo.stats -b -l ./lh.entorhinal_exvivo.label bert lh white) I was expecting the numbers to match or at-least be close; I am guessing I am using mri_label2vol incorrectly ? Could someone help me with this issue ? Thanks Mehul ___ 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.
Re: [Freesurfer] creating an entorhinal cortex volume from ?h.entorhinal_exvivo.label
Hi Bruce, Yes, the one from label2vol just hugs the white boundary Doug, I will try with the --proj frac 0 1 .1 option and let you all know. Thanks Mehul On Wed, Jun 26, 2013 at 3:58 PM, Douglas N Greve gr...@nmr.mgh.harvard.eduwrote: Oh, right. Mehul, you'll need to tell it to fill the ribbon with something like --proj frac 0 1 .1 doug On 06/26/2013 06:55 PM, Bruce Fischl wrote: won't the anatomical_stats one multiply thickness by surface area, where the label2vol will only fill voxels on the gray/white boundary, not all the way through the ribbon? Mehul: if you visualize the one from label2vol does it fill the ribbon or just hug the white boundary? On Wed, 26 Jun 2013, Douglas N Greve wrote: not necessarily. That is a very small structure, and getting a few edge voxels wrong will mess things up a lot. Can you load lh-new-erc.mgz in tkmedit and load the surfaces too and see how good it did? doug On 06/26/2013 06:49 PM, Mehul Sampat wrote: Hi Folks, I am trying to create an entorhinal cortex volume from ?h.entorhinal_exvivo.label using mri_label2vol I cd to the subject dir ($SUBJECTS_DIR/bert) and the command I use is: *mri_label2vol --label label/lh.entorhinal_exvivo.label --temp mri/nu.mgz --identity --o mri/lh-new-erc.mgz* I took the sum of the voxels in this new mask i created (lh-new-erc.mgz) and it is 215; I looked at the volume provided inlh.entorhinal_exvivo.stats and it is 925; (from the log file I believe this is created using the command: mris_anatomical_stats -mgz -f ../stats/lh.entorhinal_exvivo.stats -b -l ./lh.entorhinal_exvivo.label bert lh white) I was expecting the numbers to match orat-leastbe close; I am guessing I am using mri_label2vol incorrectly ? Could someone help me with this issue ? Thanks Mehul ___ 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 gr...@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.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.
Re: [Freesurfer] longer run time on 5.3 versus 5.2 ?
Hi Bruce, Okay. It went from ~4 hours in 5.2 to ~9 hours in 5.3; we thought it was a big change to and are trying to isolate the root cause We are running 75 cases run with 5.3 that have been run with 5.2 before; once they are done, will let you all know if this issue is seen in other cases too. Mehul ps: the cases were run on different machines but the specs of the two machines are quite similar; we are also trying to see if it is hard-ware related.. On Tue, Jun 4, 2013 at 10:37 AM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Mehul no, that's definitely not expected for it to go from 20 min to almost 5 hours! We'll investigate as that shouldn't be the case. Bruce On Tue, 4 Jun 2013, Mehul Sampat wrote: In some cases, we are seeing that, for the same subject, the run time is slower for 5.3 versus 5.2; A quick check on recon-all-status.log shows that the CA Reg Inv step is taking longer with 5.3. recon-all-status.log from data processed with 5.2: #@# CA Reg Thu Jan 17 16:03:53 PST 2013 #@# CA Reg Inv Thu Jan 17 20:17:32 PST 2013 recon-all-status.log from data processed with 5.3 #@# CA Reg Mon Jun 3 19:46:37 PDT 2013 #@# CA Reg Inv Tue Jun 4 04:52:57 PDT 2013 Is this expected ? Or in your own testing are the recon-all processing times similar for 5.2 versus 5.3 ? The recon-all.log file for both cases are attached; Thanks Mehul 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/**compliancelinehttp://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.
Re: [Freesurfer] error while running autorecon1 and autorecon2 in fs-v5.3
Hi Doug, Yes, we see the same error as in msg28047.html; here is the line from my recon-all.log: mghRead(/mnt/horizon/stage/HLC//004_151_3106223/WBS/Baseline/Freesurfer/mri/ribbon.mgz, -1): could not open file In msg28047.html, Milkos suggested that the issue could be the following line in recon-all set DoSegStats = 1 (in recon-all script in 5.2 this is line 5225; under the case -autorecon2 ) (in recon-all script in 5.3 this is line 5401; under the case -autorecon2 ) I think he is right; first i was confused why this issue was not seen with -autrecon2-cp because there is set DoSegStats = 1 for -autrecon2-cp too; (in recon-all script in 5.3 this is line 5480; under the case -autorecon2-cp) But normally we run -autorecon3 or -all before running -autorecon2-cp; so in this case the ribbon exists and the error does not show up with -autorecon2-cp (if -all or -autorecon3 is run before). Mehul On Tue, Jun 4, 2013 at 10:49 AM, Douglas N Greve gr...@nmr.mgh.harvard.eduwrote: The error in msg28047.html is that there was no ribbon.mgz. Is that the error you are receiving? http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg28047.html On 06/04/2013 01:45 PM, Mehul Sampat wrote: Hi Zeke, Thanks for looking into this; had a quick followup item to report. So for the same subject recon-all -all runs through smoothly; (it was only recon-all -autorecon1 -autorecon2 that generated the error) I had not seen the error before since in the past, my workflow was as follows: 1. run recon-all with -all 2. add control points if required 3. run recon-all with -autorecon2-cp and autorecon3 This ran through without any errors; Now to save some processing time, the modified workflow is: 1. run recon-all with -autorecon1 -autorecon2 2. add control points if required 3. run recon-all with -autorecon2-cp and autorecon3 (It was step 1 in this modified workflow that generated the issue) Mehul ps: the recon-all.log file from the error reported earlier is attached for reference (i sent it to you but forgot to send it to the mailing list earlier) On Thu, May 30, 2013 at 8:31 AM, Z K zkauf...@nmr.mgh.harvard.edu mailto:zkauf...@nmr.mgh.harvard.edu wrote: Could you provide the recon-all.log file? -Zeke On 05/30/2013 10:24 AM, Gennan Chen wrote: Does 5.3 assume recon-all -all will always run first? Since ribbon.mgz will not get created till autorecon3. In V4, that is part of autorecon2 though. Gen On Thu, May 30, 2013 at 12:24 AM, Mehul Sampat mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com wrote: ps: i believe my error is the same as the one in this thread: http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg28047.html Mehul On Thu, May 30, 2013 at 12:13 AM, Mehul Sampat mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com wrote: Hi Folks, I am running recon-all with -autorecon1 -autorecon2, -nomotioncor -notal-check -nonuintensitycor (this is with fs version 5.3) it almost runs through but toward the end, when mri_segstats is called I get the error message: .../mri/ribbon.mgz, -1): could not open file any ideas why this could be occurring ? Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu mailto:Freesurfer@nmr.mgh.harvard.edu mailto:Freesurfer@nmr.mgh.harvard.edu mailto: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 mailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu mailto
Re: [Freesurfer] error while running autorecon1 and autorecon2 in fs-v5.3
Hi Doug, Thanks; Will do so. Also I forgot to mention one thing in the previous email: for case -autorecon3, also there is: set DoSegStats = 1; (in recon-all script in 5.3 this is line 5526) would it help, if set DoSegStats = 1 was used only in -autorecon3 and not in the -autorecon2 or -autorecon2-cp ? or would this break other stuff. Mehul On Tue, Jun 4, 2013 at 11:24 AM, Douglas N Greve gr...@nmr.mgh.harvard.eduwrote: this is a bug in mri_segstats. For now, you will need to run recon-all to the end to generate the ribbon.mgz file doug On 06/04/2013 02:12 PM, Mehul Sampat wrote: Hi Doug, Yes, we see the same error as in msg28047.html; here is the line from my recon-all.log: mghRead(/mnt/horizon/stage/**HLC//004_151_3106223/WBS/** Baseline/Freesurfer/mri/**ribbon.mgz, -1): could not open file In msg28047.html, Milkos suggested that the issue could be the following line in recon-all set DoSegStats = 1 (in recon-all script in 5.2 this is line 5225; under the case -autorecon2 ) (in recon-all script in 5.3 this is line 5401; under the case -autorecon2 ) I think he is right; first i was confused why this issue was not seen with -autrecon2-cp because there is set DoSegStats = 1 for -autrecon2-cp too; (in recon-all script in 5.3 this is line 5480; under the case -autorecon2-cp) But normally we run -autorecon3 or -all before running -autorecon2-cp; so in this case the ribbon exists and the error does not show up with -autorecon2-cp (if -all or -autorecon3 is run before). Mehul On Tue, Jun 4, 2013 at 10:49 AM, Douglas N Greve gr...@nmr.mgh.harvard.edu mailto:gr...@nmr.mgh.harvard.**edugr...@nmr.mgh.harvard.edu wrote: The error in msg28047.html is that there was no ribbon.mgz. Is that the error you are receiving? http://www.mail-archive.com/**freesur...@nmr.mgh.harvard.** edu/msg28047.htmlhttp://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg28047.html On 06/04/2013 01:45 PM, Mehul Sampat wrote: Hi Zeke, Thanks for looking into this; had a quick followup item to report. So for the same subject recon-all -all runs through smoothly; (it was only recon-all -autorecon1 -autorecon2 that generated the error) I had not seen the error before since in the past, my workflow was as follows: 1. run recon-all with -all 2. add control points if required 3. run recon-all with -autorecon2-cp and autorecon3 This ran through without any errors; Now to save some processing time, the modified workflow is: 1. run recon-all with -autorecon1 -autorecon2 2. add control points if required 3. run recon-all with -autorecon2-cp and autorecon3 (It was step 1 in this modified workflow that generated the issue) Mehul ps: the recon-all.log file from the error reported earlier is attached for reference (i sent it to you but forgot to send it to the mailing list earlier) On Thu, May 30, 2013 at 8:31 AM, Z K zkauf...@nmr.mgh.harvard.edu mailto:zkauf...@nmr.mgh.**harvard.eduzkauf...@nmr.mgh.harvard.edu mailto:zkauf...@nmr.mgh.**harvard.eduzkauf...@nmr.mgh.harvard.edu mailto:zkauf...@nmr.mgh.**harvard.edu zkauf...@nmr.mgh.harvard.edu wrote: Could you provide the recon-all.log file? -Zeke On 05/30/2013 10:24 AM, Gennan Chen wrote: Does 5.3 assume recon-all -all will always run first? Since ribbon.mgz will not get created till autorecon3. In V4, that is part of autorecon2 though. Gen On Thu, May 30, 2013 at 12:24 AM, Mehul Sampat mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com wrote: ps: i believe my error is the same as the one in this thread: http://www.mail-archive.com/**freesur...@nmr.mgh.harvard.** edu/msg28047.htmlhttp://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg28047.html Mehul On Thu, May 30, 2013 at 12:13 AM, Mehul Sampat mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com mailto:mpsam...@gmail.com wrote: Hi Folks, I am running recon-all with -autorecon1 -autorecon2, -nomotioncor -notal-check -nonuintensitycor (this is with fs version 5.3
[Freesurfer] error while running autorecon1 and autorecon2 in fs-v5.3
Hi Folks, I am running recon-all with -autorecon1 -autorecon2, -nomotioncor -notal-check -nonuintensitycor (this is with fs version 5.3) it almost runs through but toward the end, when mri_segstats is called I get the error message: .../mri/ribbon.mgz, -1): could not open file any ideas why this could be occurring ? Thanks Mehul ___ 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.
Re: [Freesurfer] error while running autorecon1 and autorecon2 in fs-v5.3
ps: i believe my error is the same as the one in this thread: http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg28047.html Mehul On Thu, May 30, 2013 at 12:13 AM, Mehul Sampat mpsam...@gmail.com wrote: Hi Folks, I am running recon-all with -autorecon1 -autorecon2, -nomotioncor -notal-check -nonuintensitycor (this is with fs version 5.3) it almost runs through but toward the end, when mri_segstats is called I get the error message: .../mri/ribbon.mgz, -1): could not open file any ideas why this could be occurring ? Thanks Mehul ___ 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] question about hippocampus subfield segmentation flag
Hi Folks, I had a question about the hippocampus subfield segmentations as described here : http://surfer.nmr.mgh.harvard.edu/fswiki/HippocampalSubfieldSegmentation I am planning to use this processing outline: 1. recon-all -s subject -autorecon1 -autorecon2 2. add control points 3. recon -all -s subject -autorecon2-cp 4. recon -all -s subject -autorecon3 -hippo-subfields Is this permissible or do I need to need to introduce the -hippo-subfields flag earlier. Thanks Mehul ___ 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] question about autorecon2-cp, autorecon-pial and autorecon-wm flags
Hi Folks, I was looking at the autorecon2-cp, autorecon-pial and autorecon-wm flags in recon-all help and also here http://surfer.nmr.mgh.harvard.edu/fswiki/OtherUsefulFlags My interpretation is when I use autorecon-pial, I only need the command: recon-all --autorecon-pial -s subject While for autorecon2-cp and autorecon2-wm we need the command: recon-all -autorecon2-cp -autorecon3 -s subject recon-all -autorecon2-wm -autorecon3 -s subject since for autorecon2-cp and autorecon2-wm only a subset of stages are processed as described in recon-all help: -autorecon2-cp : process stages 12-23 (uses -f w/ mri_normalize, -keep w/ mri_seg) -autorecon2-wm : process stages 15-23 Thanks Mehul ___ 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.
Re: [Freesurfer] Process flow: ReconAllDevTable
Hi Nick I have a followup question about Scenario D: Scenario D (for pial edits): 1. recon-all -s subject -autorecon1 -autorecon2 2. edit brainmask.mgz ( http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/PialEdits) 3. recon-all -s subject -autorecon-pial I think -autorecon-pial is a subset of -autorecon3 ? Is so then could i still run -autorecon-pial without running -autorecon3? Or if i did this would i miss some of the outputs generated by -autorecon3. Also if -autorecon-pial is a subset of autorecon3, is the following scenario valid. Scenario E (for pial edits): 1. recon-all -s subject -autorecon1 -autorecon2 2. edit brainmask.mgz ( http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/PialEdits) 3. recon-all -s subject *-autorecon3 (instead of -autorecon-pial)* Thanks Mehul On Wed, Mar 13, 2013 at 12:31 PM, Mehul Sampat mpsam...@gmail.com wrote: Hi Nick, Thank you very much. I have two followup questions. Since Scenario B is permissible, I believe the following two scenarios should also be allowed. (Just wanted to confirm this:) Scenario C (for white matter edits): 1. recon-all -s subject -autorecon1 -autorecon2 2. edit wm.mgz ( http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/WhiteMatterEdits) 3. recon-all -s subject -autorecon2-wm -autorecon3 Scenario D (for pial edits): 1. recon-all -s subject -autorecon1 -autorecon2 2. edit brainmask.mgz ( http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/PialEdits) 3. recon-all -s subject -autorecon2-pial Basically, I am trying to see if I can avoid running -autorecon3 in the first run; make all the edits and then re-run recon-all with appropriate flags. Thanks Mehul On Wed, Mar 13, 2013 at 10:58 AM, Nick Schmansky ni...@nmr.mgh.harvard.edu wrote: Mehul, Scenerio B is permissible. And to address your question, the pial surface is created in the autorecon3 stage, making use of the parcellation data to refine it. I think a pial is generated during make_final_surfaces as its normal output, but its overwritten in autorecon3. Nick On Wed, 2013-03-13 at 10:52 -0700, Mehul Sampat wrote: ps: just wanted to add a clarification to my question. The two scenarios are: Scenario A: 1. recon-all -s subject -autorecon1 -autorecon2 -autorecon3 2. add control points 3. recon -all -s subject -autorecon2-cp -autorecon3 Scenario B: 1. recon-all -s subject -autorecon1 -autorecon2 2. add control points 3. recon -all -s subject -autorecon2-cp -autorecon3 If Scenario B is permissible, the advantage is that, -autorecon3 is only run once thus saving a few hours of computation. Thanks Mehul On Wed, Mar 13, 2013 at 10:35 AM, Mehul Sampat mpsam...@gmail.com wrote: Hi Folks, Based on the tutorials, we normally run full recon-all pipeline; then add control points if required and then run -autorecon2-cp and -autorecon3 again. Recently, I was looking at the process flow table: http://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable and I have two questions: 1. From this table it seems like ?h.white is created in autorecon2 and ?h.pial is created in autorecon3. However, when i run recon-all -s subj -autorecon1 -autorecon2 i see that ?h.pial is also already created. Does this mean I am interpreting the process flow table incorrectly or is there an error in the table ? 2. Also if ?h.pial and ?h.white are already created at the end of autorecon2; then can we add control points at immediately after autorecon2 ? This way we would need to run autorecon3 only once and save resources. Or am I missing something and is it that one must run -autorecon2 and -autorecon3 and then add control points and then run -autorecon2-cp and -autorecon3 again. Thanks Mehul ___ 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
[Freesurfer] Process flow: ReconAllDevTable
Hi Folks, Based on the tutorials, we normally run full recon-all pipeline; then add control points if required and then run -autorecon2-cp and -autorecon3 again. Recently, I was looking at the process flow table: http://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable and I have two questions: 1. From this table it seems like ?h.white is created in autorecon2 and ?h.pial is created in autorecon3. However, when i run recon-all -s subj -autorecon1 -autorecon2 i see that ?h.pial is also already created. Does this mean I am interpreting the process flow table incorrectly or is there an error in the table ? 2. Also if ?h.pial and ?h.white are already created at the end of autorecon2; then can we add control points at immediately after autorecon2 ? This way we would need to run autorecon3 only once and save resources. Or am I missing something and is it that one must run -autorecon2 and -autorecon3 and then add control points and then run -autorecon2-cp and -autorecon3 again. Thanks Mehul ___ 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.
Re: [Freesurfer] Process flow: ReconAllDevTable
ps: just wanted to add a clarification to my question. The two scenarios are: Scenario A: 1. recon-all -s subject -autorecon1 -autorecon2 -autorecon3 2. add control points 3. recon -all -s subject -autorecon2-cp -autorecon3 Scenario B: 1. recon-all -s subject -autorecon1 -autorecon2 2. add control points 3. recon -all -s subject -autorecon2-cp -autorecon3 If Scenario B is permissible, the advantage is that, -autorecon3 is only run once thus saving a few hours of computation. Thanks Mehul On Wed, Mar 13, 2013 at 10:35 AM, Mehul Sampat mpsam...@gmail.com wrote: Hi Folks, Based on the tutorials, we normally run full recon-all pipeline; then add control points if required and then run -autorecon2-cp and -autorecon3 again. Recently, I was looking at the process flow table: http://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable and I have two questions: 1. From this table it seems like ?h.white is created in autorecon2 and ?h.pial is created in autorecon3. However, when i run recon-all -s subj -autorecon1 -autorecon2 i see that ?h.pial is also already created. Does this mean I am interpreting the process flow table incorrectly or is there an error in the table ? 2. Also if ?h.pial and ?h.white are already created at the end of autorecon2; then can we add control points at immediately after autorecon2 ? This way we would need to run autorecon3 only once and save resources. Or am I missing something and is it that one must run -autorecon2 and -autorecon3 and then add control points and then run -autorecon2-cp and -autorecon3 again. Thanks Mehul ___ 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.
Re: [Freesurfer] Process flow: ReconAllDevTable
Hi Nick, Thank you very much. I have two followup questions. Since Scenario B is permissible, I believe the following two scenarios should also be allowed. (Just wanted to confirm this:) Scenario C (for white matter edits): 1. recon-all -s subject -autorecon1 -autorecon2 2. edit wm.mgz ( http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/WhiteMatterEdits) 3. recon-all -s subject -autorecon2-wm -autorecon3 Scenario D (for pial edits): 1. recon-all -s subject -autorecon1 -autorecon2 2. edit brainmask.mgz ( http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/PialEdits) 3. recon-all -s subject -autorecon2-pial Basically, I am trying to see if I can avoid running -autorecon3 in the first run; make all the edits and then re-run recon-all with appropriate flags. Thanks Mehul On Wed, Mar 13, 2013 at 10:58 AM, Nick Schmansky ni...@nmr.mgh.harvard.eduwrote: Mehul, Scenerio B is permissible. And to address your question, the pial surface is created in the autorecon3 stage, making use of the parcellation data to refine it. I think a pial is generated during make_final_surfaces as its normal output, but its overwritten in autorecon3. Nick On Wed, 2013-03-13 at 10:52 -0700, Mehul Sampat wrote: ps: just wanted to add a clarification to my question. The two scenarios are: Scenario A: 1. recon-all -s subject -autorecon1 -autorecon2 -autorecon3 2. add control points 3. recon -all -s subject -autorecon2-cp -autorecon3 Scenario B: 1. recon-all -s subject -autorecon1 -autorecon2 2. add control points 3. recon -all -s subject -autorecon2-cp -autorecon3 If Scenario B is permissible, the advantage is that, -autorecon3 is only run once thus saving a few hours of computation. Thanks Mehul On Wed, Mar 13, 2013 at 10:35 AM, Mehul Sampat mpsam...@gmail.com wrote: Hi Folks, Based on the tutorials, we normally run full recon-all pipeline; then add control points if required and then run -autorecon2-cp and -autorecon3 again. Recently, I was looking at the process flow table: http://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable and I have two questions: 1. From this table it seems like ?h.white is created in autorecon2 and ?h.pial is created in autorecon3. However, when i run recon-all -s subj -autorecon1 -autorecon2 i see that ?h.pial is also already created. Does this mean I am interpreting the process flow table incorrectly or is there an error in the table ? 2. Also if ?h.pial and ?h.white are already created at the end of autorecon2; then can we add control points at immediately after autorecon2 ? This way we would need to run autorecon3 only once and save resources. Or am I missing something and is it that one must run -autorecon2 and -autorecon3 and then add control points and then run -autorecon2-cp and -autorecon3 again. Thanks Mehul ___ 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] FS 5.2-beta run-times on Amazon Web services (AWS)
Hi Folks, Just wanted to share our experience with running FS 5.2-beta on Amazon Web Services (AWS). Basically, AWS has multiple instance types ( http://aws.amazon.com/ec2/instance-types/) and we were trying to figure out the most cost-effective approach. We ran two subjects through FS 5.2-beta on M1 Large Instance (m1.large) and Cluster Compute Eight Extra Large Instance (cc2.8xlarge). (same subjects run on both instance). We expected cc2.8xlarge to be faster (but it is also more expensive: $2.4 per hour; 8 cores); The run-times we got: instance-type subject start-time end-time run-time m1.large subject-1 01:05:44 UTC 2013 15:40:45 UTC 2013 *~14hr-35mins* m1.large subject-2 01:06:06 UTC 2013 15:08:45 UTC 2013 *~14hr-02mins * cc2.8xlarge subject-1 01:26:38 UTC 2013 12:30:23 UTC 2013 *~11hr-04mins* cc2.8xlarge subject-2 01:27:28 UTC 2013 12:19:08 UTC 2013 *~10hr-52mins* Although m1.large is a few hours slower, it seems to be the more cost effective option since it is $0.24 per hour (2 cores). If you have run Freesurfer on AWS, do you have a similar experience ? Any suggestions to speed up the run-times on AWS would be very helpful. Thanks Mehul ___ 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.
Re: [Freesurfer] FS 5.2-beta run-times on Amazon Web services (AWS)
Hi Bruce, No I did not specify the # of open mp threads on the recon-all cmd line. These run times were obtained by running one subject per core. for example cc2.8xlarge has 8 cores and so we ran 8 subjects at once; Thanks for the info about the # of open mp threads options; I will look into it. One other note: Bruce, Nick did you improve the memory management in 5.2 ? On our local machine we noticed we can run 6 subjects simultaneously even though we only have 12gb of ram. I thought some of the them might crash since we only have 2gb per subject but no crashes so far over 30 subjects.. Mehul On Fri, Feb 1, 2013 at 7:12 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Mehul did you specify the # of open mp threads on the recon-all cmd line? cheers Bruce On Fri, 1 Feb 2013, Mehul Sampat wrote: Hi Folks, Just wanted to share our experience with running FS 5.2-beta on Amazon Web Services (AWS). Basically, AWS has multiple instance types (http://aws.amazon.com/ec2/**instance-types/http://aws.amazon.com/ec2/instance-types/) and we were trying to figure out the most cost-effective approach. We ran two subjects through FS 5.2-beta on M1 Large Instance (m1.large) and Cluster Compute Eight Extra Large Instance (cc2.8xlarge). (same subjects run on both instance). We expected cc2.8xlarge to be faster (but it is also more expensive: $2.4 per hour; 8 cores); The run-times we got: instance-type subject start-time end-time run-time m1.large subject-1 01:05:44 UTC 2013 15:40:45 UTC 2013 ~14hr-35mins m1.large subject-2 01:06:06 UTC 2013 15:08:45 UTC 2013 ~14hr-02mins cc2.8xlarge subject-1 01:26:38 UTC 2013 12:30:23 UTC 2013 ~11hr-04mins cc2.8xlarge subject-2 01:27:28 UTC 2013 12:19:08 UTC 2013 ~10hr-52mins Although m1.large is a few hours slower, it seems to be the more cost effective option since it is $0.24 per hour (2 cores). If you have run Freesurfer on AWS, do you have a similar experience ? Any suggestions to speed up the run-times on AWS would be very helpful. Thanks Mehul 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/**compliancelinehttp://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.
Re: [Freesurfer] entorhinal cortex volume question
Hi Folks, I have a question about entorhinal cortex volume. The entorhinal cortex is listed in ?h.aparc.stats and ?h.entorhinal_exvivo.stats. In a previous post on the mailing list, it was mentioned that we should use the volume measurement from ?h.entorhinal_exvivo.stats. We would like to make manual corrections to the entorhinal cortex segmentation (when required). For this task I can create a entorhinal volume from ?h.entorhinal_exvivo.label using mri_label2vol This entorhinal volume will be different from the one in aparc+aseg.mgz My question is: Has anyone published any protocols to somehow combine the two entorhinal volumes ? Or is recommend to take the take the entorhinal volume created from ?h.entorhinal_exvivo.stats. and overlay it on norm.mgz and edit manually (when required). One limitation of this approach is that it would be hard to integrate this entorhinal volume back into aparc+aseg.mgz... Thanks Mehul On Wed, Dec 7, 2011 at 10:50 AM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Catherine I guess I would recommend the ex vivo one since it is based explicitly on architecontics and not on guessing locations from folds. cheers Bruce On Wed, 7 Dec 2011, Cat Chong wrote: Hello experts, I am very new to freesurfer so please excuse my basic question: We want to get measurements of entorhinal cortex volumes on a group of people. I found these results listed in ?h.aparc.stats. I also noticed another ?h.entorhinal_exvivo.stats, with different results. Which should I use for a group analysis of entorhinal volume and surface area, and why are the results very different? very best regards, catherine ___ 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.
Re: [Freesurfer] entorhinal cortex volume question
Thanks for quick response Bruce. We will use the volume from ex-vivo stats.. Regarding the need to integrate the ex vivo one back into aparc+aseg. We noticed the entorhinal volume in aparc is larger than that in exvivo.stats. So do the extra entorhinal voxels in aparc need to be re-labelled as unkown or as cortical-gray-matter? We plan to make manual correction on aparc (when required): So would it be a good approach to use the entorhinal labels from exvivo.stats to guide the corrections on aparc ? In this case, the question is, what should the end-user re-label the extra entorhinal voxels as ? Thanks Mehul On Tue, Jan 29, 2013 at 6:02 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Mehul the entorhinal label in the aparc was drawn by Rahul based on gyral landmarks. The one in the _exvivo.stats is based on being able to see the layer II islands in ex vivo MRI and hence we believe it is more accurate. Not sure why you need to integrate the ex vivo one back into the aparc+aseg, can you clarify? cheers Bruce On Tue, 29 Jan 2013, Mehul Sampat wrote: Hi Folks, I have a question about entorhinal cortex volume. The entorhinal cortex is listed in ?h.aparc.stats and ?h.entorhinal_exvivo.stats. In a previous post on the mailing list, it was mentioned that we should use the volume measurement from ?h.entorhinal_exvivo.stats. We would like to make manual corrections to the entorhinal cortex segmentation (when required). For this task I can create a entorhinal volume from ?h.entorhinal_exvivo.label using mri_label2vol This entorhinal volume will be different from the one in aparc+aseg.mgz My question is: Has anyone published any protocols to somehow combine the two entorhinal volumes ? Or is recommend to take the take the entorhinal volume created from ?h.entorhinal_exvivo.stats. and overlay it on norm.mgz and edit manually (when required). One limitation of this approach is that it would be hard to integrate this entorhinal volume back into aparc+aseg.mgz... Thanks Mehul On Wed, Dec 7, 2011 at 10:50 AM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: Hi Catherine I guess I would recommend the ex vivo one since it is based explicitly on architecontics and not on guessing locations from folds. cheers Bruce On Wed, 7 Dec 2011, Cat Chong wrote: Hello experts, I am very new to freesurfer so please excuse my basic question: We want to get measurements of entorhinal cortex volumes on a group of people. I found these results listed in ?h.aparc.stats. I also noticed another ?h.entorhinal_exvivo.stats, with different results. Which should I use for a group analysis of entorhinal volume and surface area, and why are the results very different? very best regards, catherine __**_ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.**edu/mailman/listinfo/**freesurferhttps://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/**compliancelinehttp://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] Freesurfer 5.2 release date?
Hi Folks, I just wanted to inquire about the estimated release date for Freesurfer 5.2 ? Is there a beta version we could try out ? Thanks Mehul ___ 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.
Re: [Freesurfer] expert option error in recon-all
Hi James, I had the same issue. The following workaround was mentioned on the FS mailing list: http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg20335.html It works for me. Mehul On Fri, Aug 3, 2012 at 10:11 AM, Douglas N Greve gr...@nmr.mgh.harvard.eduwrote: Delete that dir and run recon-all with -debug as the first option. Pipe the output to a file, eg, recon-all -debug ... | tee recon-all.log And send me recon-all.log. Note it will likely be very big! doug On 08/03/2012 12:52 PM, james pardon wrote: works fine. Even I can cp expert.opts to scripts folder on my own. Sorry forgot to reply all at first On Fri, Aug 3, 2012 at 9:13 PM, Douglas N Greve gr...@nmr.mgh.harvard.edu mailto:gr...@nmr.mgh.harvard.edu wrote: So /home/dir/subject/subname/scripts does not exist? If so that is wrong. What happens if you mkdir /home/dir/subject/subname/scripts doug On 08/03/2012 12:09 PM, james pardon wrote: Nick or Doug could you help us here? On Fri, Aug 3, 2012 at 8:05 PM, Bruce Fischl fis...@nmr.mgh.harvard.edu mailto:fis...@nmr.mgh.harvard.edu mailto:fis...@nmr.mgh.harvard.edu mailto:fis...@nmr.mgh.harvard.edu wrote: Hi James can you post to the list so others can answer? I'm not sure what's going on, but Nick or Doug might Bruce On Fri, 3 Aug 2012, james pardon wrote: Hi Bruce, subject's folder is empty, there is no scripts folder, or anything else. Again if I omit -expert everything works fine. Thanks On Fri, Aug 3, 2012 at 7:29 PM, Bruce Fischl fis...@nmr.mgh.harvard.edu mailto:fis...@nmr.mgh.harvard.edu mailto:fis...@nmr.mgh.harvard.edu mailto:fis...@nmr.mgh.harvard.edu wrote: Hi James what happens if you run: ls -l /home/dir/subject/subname/scripts/expert-options touch /home/dir/subject/subname/scripts/expert-options Bruce On Fri, 3 Aug 2012, james pardon wrote: Dear all, I'm trying to pass the following flags for mri_normalize through -expert flag in recon-all, with a file, named expert.opts with the following single line: mri_normalize -b 20 -n 5 my command line is: recon-all -i inputfile -subject subname -all -expert pathname/expert.opts I receive the following error: cp: cannot create regular file: '/home/dir/subject/subname/scripts/expert-options' The directory of the subject has been created, but it is completely empty. At first glance I thought it might be some sort of permission problems, however I have rwx permission for all the paths (even running under the root username the same error comes up). Secondly, I checked my disk quota, it is completely fine. Please note that, if I do not pass -expert flag recon-all works completely fine. Any input is much appreciated. Regards, James 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 gr...@nmr.mgh.harvard.edu mailto:gr...@nmr.mgh.harvard.edu Phone Number: 617-724-2358 tel:617-724-2358 Fax: 617-726-7422
Re: [Freesurfer] WM/Pial surfs Problem with SPGRs from 3T GE
Hi Thomas, The following message posted by Michael Harms http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg20991.html has some very useful options. I recently noticed that they help us with a very similar problem... Also, I recall that for FS v5.0 there was a flag -nuintensitycor-3T for optimal parameters for nu_correct for 3T scans (as described in release notes for v5.0.) Bruce, the -nuintensitycor-3T is still available in v5.1 right? If yes, then we could combine the options the two set of options and run recon-all with -b 20 -n 5 and -nuintensitycor-3T for the 3T scans? Would this be reasonable or does this combination not recommended ? Thanks Mehul On Tue, Jul 24, 2012 at 12:47 AM, Thomas Fink mr.thomas.f...@googlemail.com wrote: Hey Experts, As you can see from the picture, the autorecon does not work properly (for WM/Pial surfs) with my files: Especially in the temporal lobes. I am working with the SPGR files of a 3T GE Machine. Is there any way to optimize the autorecon results for SPGRs? (Except the lavish application of CPs.) Best regards Thomas On Mon, Jul 23, 2012 at 6:00 PM, freesurfer-requ...@nmr.mgh.harvard.eduwrote: Send Freesurfer mailing list submissions to freesurfer@nmr.mgh.harvard.edu To subscribe or unsubscribe via the World Wide Web, visit https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer or, via email, send a message with subject or body 'help' to freesurfer-requ...@nmr.mgh.harvard.edu You can reach the person managing the list at freesurfer-ow...@nmr.mgh.harvard.edu When replying, please edit your Subject line so it is more specific than Re: Contents of Freesurfer digest... Today's Topics: 1. Re: Questions about correction over 2 hemispheres and MCC (Reem Jan) 2. More than two time-points in longitudinal base = Memory allocation error? (Liz Bowman) 3. Re: Talairach registration (Mojdeh Zamyadi) 4. Subcortical segmentations (Jordan Pierce) -- Message: 1 Date: Sun, 22 Jul 2012 21:12:44 + From: Reem Jan r@auckland.ac.nz Subject: Re: [Freesurfer] Questions about correction over 2 hemispheres and MCC To: Watson, Christopher christopher.wat...@childrens.harvard.edu Cc: Freesurfer Mailinglist freesurfer@nmr.mgh.harvard.edu Message-ID: 3375b664d4a0834fb7d0c15d590a9ef11825d...@fmhs-mbx1.fmhs.auckland.ac.nz Content-Type: text/plain; charset=iso-8859-1 Hi Bruce, J?rgen and Chris Thank you for elaborating further. I've gone with a Bonferroni correction as I had a limited number of tests (a-priori hypothesis). But just out of interest Chris, you mentioned you found correlations between subcortical structures, do you have a reference I could look at? The tool that J?rgen suggested requires we enter a correlation coefficient, so I wondered if I could use published values or whether I'd need to calculate my own (if so, could you please recommend a way of doing that in Freesurfer or FSL?) Many thanks in advance Kind regards Reem -Original Message- From: freesurfer-boun...@nmr.mgh.harvard.edu [mailto: freesurfer-boun...@nmr.mgh.harvard.edu] On Behalf Of J?rgen H?nggi Sent: Sunday, 22 July 2012 7:47 p.m. To: Watson, Christopher Cc: Freesurfer Mailinglist Subject: Re: [Freesurfer] Questions about correction over 2 hemispheres and MCC Hi Chris Yes there are such tools. Bonferroni oder Sidak test that take into account the correlated measures. Here you can find these tools http://www.quantitativeskills.com/sisa/calculations/bonfer.htm Cheers J?rgen On [DATE], Watson, Christopher [ADDRESS] wrote: What about, for example, the correlations I've seen in a cohort of subjects. In 158 subjects aged 10-19 (both controls and patients), the correlation between L R thalamus is 0.91, and the correlations between L R of caudate, putamen, pallidum, hippocampus, and amygdala were all 0.75 or higher. I would think that a Bonferroni correction would be incredibly conservative and, in my opinion, just plain wrong because true significant diff's would be missed. Is there any principled way of dealing with multiple tests that are correlated? Thanks, Chris From: Bruce Fischl [fis...@nmr.mgh.harvard.edu] Sent: Saturday, July 21, 2012 12:01 PM To: Watson, Christopher Cc: Douglas N Greve; freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Questions about correction over 2 hemispheres and MCC Hi Chris, bonferroni will be overly conservative in that case, but we rarely really know the true covariance structure of the data, so we would rather err on the side of being conservative. cheers Bruce On Fri, 20 Jul 2012, Watson, Christopher wrote: Hi Doug et al, The 2nd question is something I've wondered about. Doesn't a
Re: [Freesurfer] building a pipeline using mri_robust_register and other FS functions
Thanks Bruce and Martin. Martin, I have a followup question: Let's say I run mri_robust_register with --iscale flag and also use --halfmov and --halfdst flags to create two outputs hm.mgz and hd.mgz 1) You mentioned that --iscale adjusts the intensities of both input images to better match. Are the image intensities of the outputs hm.mgz and hd.mgz also adjusted to match each other ? Also here is brief description (in case someone else is looking for something similar) Basically, I have 2 scans of the same subject and I need three steps: (1) skull-stripping, (2) intensity normalization between time-points and (3) registration of time-points; (following your recommendation, I put skull removal before registration) For the intensity normalization, ideally, a good option would be to do a histogram matching between the two time-points. However, I am guessing that the mri_normalize would be a good first approximation/substitute to the histogram matching step. (I did see mri_histo_eq function but I am not sure if I should use that instead of mri_normalize.) Alternatively, Bruce suggested I could use the FS longitudinal pipeline and compare the norm.mgz images for each time-point from the longitudinal data. I like this approach better since the longitudinal stream maps all the timepoints to the template space and I can look for changes in the template space. Thanks Mehul On Wed, Jan 25, 2012 at 1:45 PM, Martin Reuter mreu...@nmr.mgh.harvard.eduwrote: Hi Mehul, - if lesions show large changes, normalization might be dangerous - mri_robust_register has a flag --iscale for global intensity adjustment (a global scaling parameter that adjusts the intensity images of both inputs to better match) - mri_normalize, normalized the white matter to be around 110 is that what you want? - usually registration will be more accurate if images are skull stripped. Best, Martin On Tue, 2012-01-24 at 09:39 -0800, Mehul Sampat wrote: Hi Folks, We have subjects with high lesion load which changes significantly over time. I want to use FS functions to build a pipeline for comparing lesion changes in two time-points of the same-subject. I am thinking of using the following steps; 1) Use mri_normalize to normalize the two time-points. 2) Use mri_robust_register to register two time-points of the same subject to half-way space. 3) Use mri_skull_strip 4) Use subtraction imaging or some other techniques to look for lesion changes. My questions are: 1) I think I need mri_normalize since the output from mri_robust_register is not intensity normalized ? 2) Instead of the first three steps, I could also do the following: Run autorecon1 for both timepoints and then run mri_robust_register on the skull stripped images Does it matter if we run mri_robust_register before or after skull stripping ? Thanks Mehul ___ 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] building a pipeline using mri_robust_register and other FS functions
Hi Folks, We have subjects with high lesion load which changes significantly over time. I want to use FS functions to build a pipeline for comparing lesion changes in two time-points of the same-subject. I am thinking of using the following steps; 1) Use mri_normalize to normalize the two time-points. 2) Use mri_robust_register to register two time-points of the same subject to half-way space. 3) Use mri_skull_strip 4) Use subtraction imaging or some other techniques to look for lesion changes. My questions are: 1) I think I need mri_normalize since the output from mri_robust_register is not intensity normalized ? 2) Instead of the first three steps, I could also do the following: Run autorecon1 for both timepoints and then run mri_robust_register on the skull stripped images Does it matter if we run mri_robust_register before or after skull stripping ? Thanks Mehul ___ 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.
Re: [Freesurfer] temporal lobe cut off
Hi Clare, I had that same error a few months ago. (I had the error in 1 out of 500+cases). Someone on the list had suggested the following solution: Alternatively you can try rerunning the skullstrip step, sometimes this works as well. recon-all -skullstrip -autorecon2 -autorecon3 -s [SubjID] This had worked for me. I guess you can add the expert options to the above command and see if it works for you. Good luck Mehul On Thu, Jan 12, 2012 at 9:39 AM, Gibbard, Clare c.gibb...@ucl.ac.uk wrote: Hi Bruce, In order to solve the temporal lobe issue in my scans I ran recon-all again from scratch using an expert option to run the mri_normalize command with the -b 20 and -n5 options (as per Michael Harms' email to the mailing list): recon-all -all -subjid subject_1010 -expert /home/clare/freesurfer/subjects/subject_1010/scripts/expert.opts It seemed to be running well, but exited early with the following error message: ERROR: _FindFacePath: could not find path! I've seen that other people have had this problem, but can't see the solution. I have started uploading my subject to your file exchange (the clarefilepatherror.tar file), but it seems to be taking some time. Let me know if you don't get it. I've attached the log to this email if that helps as well. I would really appreciate some advice on fixing this. Thank you. Best wishes, Clare From: Gibbard, Clare Sent: 10 January 2012 15:38 To: Bruce Fischl Cc: freesurfer@nmr.mgh.harvard.edu; Gibbard, Clare Subject: RE: [Freesurfer] temporal lobe cut off Hi Bruce, Thank you for your quick reply. I installed v5.1.0 from the website and that name is what comes up when I open a terminal window. However, I tried typing recon-all -version and got $Id: recon-all,v 1.379.2.17 2011/05/20 22:48:18 nicks Exp $. So I'm not sure if that means the version is different from what I thought. I tried to upload 3 directories to the FTP in /transfer/incoming/claregibbard.tar. subject_1010_orig is the subject before adding control points. subject_1010_cpleft is after I added control points to the left temporal lobe. subject_1010_cpleftandright is when I tried adding control points to both the left and right temporal lobe. However, I can't tell if they have uploaded or not (it was taking a long time, but when I aborted the upload it said File receive OK). If you need me to, I can try sending you fewer files. I hope that helps. Let me know if you need any more information. Best wishes, Clare From: Bruce Fischl [fis...@nmr.mgh.harvard.edu] Sent: 10 January 2012 13:40 To: Gibbard, Clare Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] temporal lobe cut off Hi Claire, hmmm, that is odd. What version are you running? If you upload the subject before and after we'll take a look Bruce On Tue, 10 Jan 2012, Gibbard, Clare wrote: Hi, I ran recon-all -all -subjid subjectx on around 40 scans. The registrations and skull strips look fine. However, the white matter and pial surface segmentations miss out the left anterior temporal lobe in most subjects. I tried adding in control points to the left temporal lobe and running recon-all -autorecon2-cp -autorecon3 -subjid subjectx. This helped to fill in the missing areas, but new parts of the right temporal lobe were cut out. I then tried adding in control points to both the left and right temporal lobes, but again new regions of the left temporal lobe were cut out. I think it must be an intensity issue, but I can't see how to fix it. Are you able to give me a work flow to sort this? Here are some screenshots: [moz-screenshot.png] [moz-screenshot-1.png] Before control points added: [IMAGE] After control points added to left and right temporal lobes: [IMAGE] Thank you. Best wishes, Clare Gibbard 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
Re: [Freesurfer] gcut memory requirements
Hi Michael, I had the same issue a few months ago. In my experience 4GB was sufficient. If there are constraints on the memory, I think 3 GB should also work. I also had the same issue with -gcut and FS 5.0. Back then I had tried to figure out what was the minimum memory required since I had some memory constraints. I was running jobs on cluster where each node had 16 cores and 32 GB total memory (that could be divided between the 16 cores). At most I was able to run 12 recons on each node. So the minimum memory required was ~ 2.7 GB (32 GB/12). Mehul On Sat, Nov 19, 2011 at 12:46 PM, Michael Harms mha...@conte.wustl.eduwrote: Hello, Approximately how much memory does one need when using the -gcut option with FS 5.1? I had submitted 30+ recons to a cluster that alloted a default of 2.5 GB for processes submitted under this particular queue, and all of them aborted with a bad_alloc error at the gcut stage of skull stripping. Should 4 GB be sufficient? thanks, -MH -- Michael Harms, Ph.D. Conte Center for the Neuroscience of Mental Disorders Washington University School of Medicine Department of Psychiatry, Box 8134 Renard Hospital, Room 6604 Tel: 314-747-6173 660 South Euclid Ave.Fax: 314-747-2182 St. Louis, MO 63110 Email: mha...@wustl.edu ___ 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.
Re: [Freesurfer] Topology correction - segmentation fault
Hi Bruce and Louis, Thank you for the suggestions. I will try them out. Just fyi. For our dataset this error is very rare. So far it occured only in 1 out of 525+ scans processed with FS 5.1. Thank you, Mehul On Tue, Nov 1, 2011 at 6:50 AM, Louis Nicholas Vinke vi...@nmr.mgh.harvard.edu wrote: Hi Mehul, Alternatively you can try rerunning the skullstrip step, sometimes this works as well. recon-all -skullstrip -autorecon2 -autorecon3 -s [SubjID] -Louis On Tue, 1 Nov 2011, Bruce Fischl wrote: Hi Mehul this is a bug in the topology correction that we haven't been able to track down, mainly because the person who wrote this piece is long gone. If you edit the wm to remove a defect it usually goes away Bruce On Mon, 31 Oct 2011, Mehul Sampat wrote: Hi Folks I am using Freesurfer 5.1 and I get the following error: ERROR: _FindFacePath: could not find path! Is this is a memory related issue ? The same case went through smoothly with Freesurfer 5.0. Also I have not seen this error over a large number of subjects and I thought it might be system related, but even when I re-run it, I get the same error. From the recon-all.error file, the last command is mris_topo_fixer -mgz -warnings -seed 1234 ms0709_01 rh Here is a small snippet from the recon-all.log file (just before the error) Correcting Topology of defect 28 with euler number -1 (1 loops) computing statistics for defect 28: 288 vertices location: [ (106,175,134) - average intensity = 89.717 ] -gray ( 85.29 , 6.61 ) -white ( 95.96 , 3.49 ) -gray ( 83.95 , 18.89 ) -white ( 98.95 , 17.00 ) -intensity (85.286896 [log = -2.118856 ]- 95.961853 [log = -2.118159 ] ) -curv (k1=-0.131 (0.556) , r1 = 7.623 | k2=-0.051 (0.131), r2 = 19.605 ) -curv (k1=-0.105 (0.609) , r1 = 9.520 | k2=-0.023 (0.212), r2 = 42.735 ) max face = 548(548) - loop = 1 (1) - ntries = [83,229] BEST FITNESS (o)is -6.78736 mri =0.000 curv = 1.594 unmri = 0.410 ( f=0.00 , v=0.00 , c=1.59 , q= 3.19 ) ( f=0.00 , v=0.00 , c=1.59 , q= 3.19 ) BEST FITNESS (M) is -6.40381 mri =0.000 curv = 2.115 unmri = -2.054 ( f=0.00 , v=0.00 , c=2.11 , q= 4.23 ) ( f=0.00 , v=0.00 , c=2.11 , q= 4.23 ) ERROR: _FindFacePath: could not find path! Thanks Mehul On Wed, Oct 26, 2011 at 6:03 PM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: Hi you probably ran out of memory. That's a giant defect so something is badly wrong. Check to make sure that the hemis are separated and that the skull and cerebellum aren't attached Bruce On Wed, 26 Oct 2011, Sindhuja Tirumalai Govindarajan wrote: Hi all, I got the following error while running recon-all on a patient dataset. Kindly let me know how to proceed. Thanks! Sindhuja Correction of the Topology Finding true center and radius of Spherical Surface...done Surface centered at (0,0,0) with radius 100.0 in 13 iterations marking ambiguous vertices... 129422 ambiguous faces found in tessellation segmenting defects... 51 defects found, arbitrating ambiguous regions... analyzing neighboring defects... -merging segment 1 into 0 -merging segment 2 into 0 -merging segment 8 into 0 -merging segment 5 into 7 -merging segment 28 into 12 -merging segment 35 into 31 -merging segment 47 into 43 44 defects to be corrected 0 vertices coincident reading input surface /autofs/cluster/mscat/users/**caterina/3T/recons-04082011/** MS_pt20_recon/surf/ lh.qsphere.nofix... reading brain volume from brain... reading wm segmentation from wm... Computing Initial Surface Statistics -face loglikelihood: -9.5482 (-4.7741) -vertex loglikelihood: -6.9824 (-3.4912) -normal dot loglikelihood: -3.5307 (-3.5307) -quad curv loglikelihood: -5.9945 (-2.9973) Total Loglikelihood : -26.0558 CORRECTING DEFECT 0 (vertices=50712, convex hull=8198) Segmentation fault recon-all -s MS_pt20_recon exited with ERRORS at Wed Oct 26 18:49:31 EDT 2011 __**_ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.**edu/mailman/listinfo/**freesurferhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer __**_ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.**edu/mailman/listinfo/**freesurferhttps://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
Re: [Freesurfer] Topology correction - segmentation fault
Hi Folks I am using Freesurfer 5.1 and I get the following error: ERROR: _FindFacePath: could not find path! Is this is a memory related issue ? The same case went through smoothly with Freesurfer 5.0. Also I have not seen this error over a large number of subjects and I thought it might be system related, but even when I re-run it, I get the same error. From the recon-all.error file, the last command is mris_topo_fixer -mgz -warnings -seed 1234 ms0709_01 rh Here is a small snippet from the recon-all.log file (just before the error) Correcting Topology of defect 28 with euler number -1 (1 loops) computing statistics for defect 28: 288 vertices location: [ (106,175,134) - average intensity = 89.717 ] -gray ( 85.29 , 6.61 ) -white ( 95.96 , 3.49 ) -gray ( 83.95 , 18.89 ) -white ( 98.95 , 17.00 ) -intensity (85.286896 [log = -2.118856 ]- 95.961853 [log = -2.118159 ]) -curv (k1=-0.131 (0.556) , r1 = 7.623 | k2=-0.051 (0.131), r2 = 19.605 ) -curv (k1=-0.105 (0.609) , r1 = 9.520 | k2=-0.023 (0.212), r2 = 42.735 ) max face = 548(548) - loop = 1 (1) - ntries = [83,229] BEST FITNESS (o)is -6.78736 mri =0.000 curv = 1.594 unmri = 0.410 ( f=0.00 , v=0.00 , c=1.59 , q= 3.19 ) ( f=0.00 , v=0.00 , c=1.59 , q= 3.19 ) BEST FITNESS (M) is -6.40381 mri =0.000 curv = 2.115 unmri = -2.054 ( f=0.00 , v=0.00 , c=2.11 , q= 4.23 ) ( f=0.00 , v=0.00 , c=2.11 , q= 4.23 ) ERROR: _FindFacePath: could not find path! Thanks Mehul On Wed, Oct 26, 2011 at 6:03 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi you probably ran out of memory. That's a giant defect so something is badly wrong. Check to make sure that the hemis are separated and that the skull and cerebellum aren't attached Bruce On Wed, 26 Oct 2011, Sindhuja Tirumalai Govindarajan wrote: Hi all, I got the following error while running recon-all on a patient dataset. Kindly let me know how to proceed. Thanks! Sindhuja Correction of the Topology Finding true center and radius of Spherical Surface...done Surface centered at (0,0,0) with radius 100.0 in 13 iterations marking ambiguous vertices... 129422 ambiguous faces found in tessellation segmenting defects... 51 defects found, arbitrating ambiguous regions... analyzing neighboring defects... -merging segment 1 into 0 -merging segment 2 into 0 -merging segment 8 into 0 -merging segment 5 into 7 -merging segment 28 into 12 -merging segment 35 into 31 -merging segment 47 into 43 44 defects to be corrected 0 vertices coincident reading input surface /autofs/cluster/mscat/users/caterina/3T/recons-04082011/MS_pt20_recon/surf/lh.qsphere.nofix... reading brain volume from brain... reading wm segmentation from wm... Computing Initial Surface Statistics -face loglikelihood: -9.5482 (-4.7741) -vertex loglikelihood: -6.9824 (-3.4912) -normal dot loglikelihood: -3.5307 (-3.5307) -quad curv loglikelihood: -5.9945 (-2.9973) Total Loglikelihood : -26.0558 CORRECTING DEFECT 0 (vertices=50712, convex hull=8198) Segmentation fault recon-all -s MS_pt20_recon exited with ERRORS at Wed Oct 26 18:49:31 EDT 2011 ___ 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 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.
Re: [Freesurfer] I/O performance mitigation
Hi Bruce and others, My question is related to the staggering of FS jobs (see thread below). Could you tell me by how much time you stagger the freesurfer on a cluster ? I am using FS 5.1 on a cluster which uses sun grid engine and I have the following error when I try to submit a large number of jobs ..I do not see it for small job batches and I think it might be related to staggering FS jobs but I am not sure. the error message from recon-all.log: *nu_estimate_np_and_em: crashed while running spline_smooth (termination status=11) nu_correct: crashed while running nu_estimate_np_and_em (termination status=65280) * the message in recon-all.error *PWD /work/01523/msampat/**freesurfer-5.1/subjects/**ms0880_01/mri CMD mri_nu_correct.mni --i orig.mgz --o nu.mgz --uchar transforms/talairach.xfm --proto-iters 1000 --distance 50 --n 1 * First i thought, it was a memory issue but the sun grid engine is supposed to assign 4gb per core, so i thought it was enough memory to run each case. I am investigating if the memory is not allocated correctly.. If anyone knows what this error is related to, could you please let me know ? Thanks Mehul On Fri, Sep 3, 2010 at 10:25 AM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi David, I'm surprised you didn't have to do this in the past. We always space our jobs out. Glad there's an easy workaround cheers Bruce On Fri, 3 Sep 2010, David Mischel wrote: We took the suggestion of staggering the launch of Freesurfer 5.0 recon-all jobs. The attached Word doc (I don't know how to contribute this information other than attaching the image and text using Word) shows a load graph on our file server. When 20 FS jobs began at once (all processing servers using a single file server) the load on the file server bulged up. When we spaced out the launch of each job by 15 seconds the load hardly budged. We have not had to do this in the past with earlier versions of Freesurfer but this is an obvious work around to the problem we encountered. david David Mischel Manager of IT Center for Imaging of Neurodegenerative Diseases (CIND) http://www.cind.research.va.gov/ http://www.cind.research.va.gov/ VA Medical Center 4150 Clement Street, 114M San Francisco, CA 94121 voice: 415-221-4810 x3864 fax: 415-668-2864 ___ 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.
Re: [Freesurfer] I/O performance mitigation
Thanks for letting me know. On my side, my first guess is for some reason memory is not being allocated correctly by the the script I use (this mostly uses standard sun grid engine flags) I am trying to verify this with sysadmins. Just fyi. I am using the cluster on www.teragrid.org and one of the sysadmins there looked at the error and he thinks the MCSRCH routine is the cause of this error. I don't know what that is; but he suggested the following action: I found this reply from the developer of the routine that fails (MCSRCH) regarding this very same error: http://code.google.com/p/mitlm/issues/detail?id=10 Can you try that and see if it works? I will let you know if we find a solution. Thanks Mehul On Thu, May 26, 2011 at 1:11 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Mehul, we don't sorry. We are trying to replicate it here. The fact that it's in an MNI tool and not one of our's makes it harder to track down. The delay between jobs for us is driven by how robust and rapid your storage is. Certainly a minute or two should be enough I would think, altough we launch much more rapidly than that typically Bruce On Thu, 26 May 2011, Mehul Sampat wrote: Hi Bruce and others, My question is related to the staggering of FS jobs (see thread below). Could you tell me by how much time you stagger the freesurfer on a cluster ? I am using FS 5.1 on a cluster which uses sun grid engine and I have the following error when I try to submit a large number of jobs ..I do not see it for small job batches and I think it might be related to staggering FS jobs but I am not sure. the error message from recon-all.log: *nu_estimate_np_and_em: crashed while running spline_smooth (termination status=11) nu_correct: crashed while running nu_estimate_np_and_em (termination status=65280) * the message in recon-all.error *PWD /work/01523/msampat/**freesurfer-5.1/subjects/**ms0880_01/mri CMD mri_nu_correct.mni --i orig.mgz --o nu.mgz --uchar transforms/talairach.xfm --proto-iters 1000 --distance 50 --n 1 * First i thought, it was a memory issue but the sun grid engine is supposed to assign 4gb per core, so i thought it was enough memory to run each case. I am investigating if the memory is not allocated correctly.. If anyone knows what this error is related to, could you please let me know ? Thanks Mehul On Fri, Sep 3, 2010 at 10:25 AM, Bruce Fischl fis...@nmr.mgh.harvard.edu wrote: Hi David, I'm surprised you didn't have to do this in the past. We always space our jobs out. Glad there's an easy workaround cheers Bruce On Fri, 3 Sep 2010, David Mischel wrote: We took the suggestion of staggering the launch of Freesurfer 5.0 recon-all jobs. The attached Word doc (I don't know how to contribute this information other than attaching the image and text using Word) shows a load graph on our file server. When 20 FS jobs began at once (all processing servers using a single file server) the load on the file server bulged up. When we spaced out the launch of each job by 15 seconds the load hardly budged. We have not had to do this in the past with earlier versions of Freesurfer but this is an obvious work around to the problem we encountered. david David Mischel Manager of IT Center for Imaging of Neurodegenerative Diseases (CIND) http://www.cind.research.va.gov/ http://www.cind.research.va.gov/ VA Medical Center 4150 Clement Street, 114M San Francisco, CA 94121 voice: 415-221-4810 x3864 fax: 415-668-2864 ___ 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.
Re: [Freesurfer] FS 5.1
Hi Bruce, We have two studies being conducted (one cross-sectional and one longitudinal) Some of the subjects in have a lot of atrophy and have large ventircles. For the cross-sectional analysis stream are there many differences between 4.5 and the forthcoming 5.1 ? Or would it be fine to use 4.5 for cross-sectional analysis ? (I understand that for the longitudinal study, I should wait for 5.1 and process all the timepoints with 5.1) Thanks Mehul On Thu, Jan 20, 2011 at 10:36 AM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Derin we are tracking down some issues with it and hope to have it out within a month cheers Bruce On Thu, 20 Jan 2011, Derin Cobia wrote: All, Just curious about the current estimated timeline for version 5.1, trying to schedule some projects around it. Thanks! -Derin ___ 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 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] freeview and control points
Hi Folks, We used to use tkmedit to add control points and recently started using freeview. I have a few basic questions about saving controls points created with freeview. 1. I noticed that in tkmedit edit controls points were saved as control.dat in the subject_dir/tmp (e.g. bert/tmp). When I use freeview I assume I should also save the control points in the same location ? 2. If so, is there a way to configure freeview so that it saves it to subject_dir/tmp ? (currently the default location where freeview saves the control points in the current subject directory.) 3. Also I believe tkmedit saved the control points as control.dat but for freeview one has to specify the name manually. Is there a command line argument to set the name of the file as control.dat by default ? Thanks Mehul ___ 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.
Re: [Freesurfer] mri_gcut killed
Hi Bruce and Nick, I think mri_gcut seems to require a lot of memory and might be causing the crash ? I was running some subjects on a node (16 cpu 32 GB RAM ; thus 2GB RAM per cpu) and some of the subjects crashed for me at the mri_gcut step. (a few lines from recon-all.log are shown below) mri_gcut -110 -mult brainmask.auto.mgz T1.mgz brainmask.auto.mgz terminate called after throwing an instance of 'std::bad_alloc' what(): St9bad_alloc reading mask... use voxels with intensity 110 as WM mask threshold set to: 110.00*0.36=39.60 calculating weights... doing mincut... Abort - They worked if I assigned 4gb per cpu (Also i found i could run atmost 12 subjects with 32GB RAM).. maybe Michelle has a similar issue ? Mehul On Mon, Jan 24, 2011 at 1:56 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: is anything else running on the machine? Maybe some other process was taking a bunch of memory while the gcut was running? On Mon, 24 Jan 2011, Nick Schmansky wrote: confirm that both T1.mgz and brainmask.auto.mgz a decent images (via tkmedit). if they are good, then the gcut stage can be skipped by adding the -no-gcut flag to the end of the recon-all stream, but i have never seen this particular error occur, so i wonder about the quality of the input. if those two files look fine, and gcut still gets killed (which is odd), then you can send me those files via our file drop and i can try to replicate. n. On Mon, 2011-01-24 at 15:01 -0500, Michelle Umali wrote: Dear All, recon-all was killed during mri_gcut. Here is the error: mri_gcut -110 -mult brainmask.auto.mgz T1.mgz brainmask.auto.mgz Killed. How does one deal with this? Thanks in advance for any help. Michelle ___ 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 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] changing tkmedit keyboard shortcuts
Hi Folks, In tkmedit, to toggle between the main and aux volumes, one uses ctrl-1 and ctrl-2. I would like to change this to just 1 and 2 Is it possible to do so using tkmedit scripting ? If yes, could you tell me which function argument I should change ? I looked at the tkmedit scripting help page ( https://surfer.nmr.mgh.harvard.edu/fswiki/TkMeditGuide/TkMeditReference/TkMeditScripting ) but i dont know which function argument to change. Thanks Mehul ___ 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] question about using lta file from mri_robust_register
Hi Folks, I am trying to troubleshoot an issue and I have not been able to find a solution yet. Here is what I am trying to do: 1. I used mri_robust_register to register a t1-weighted image to a t2-image. mri_robust_register works very well and i also got an lta file (t1-to-t2.lta) 2. now i also have the wm-mask from the freesurfer output. this was created using: mri_binarize --i aparc+aseg.mgz --wm --o t1-wm-mask-fs.mgz 3. Then I take the wm-mask to native space using: mri_convert -rl orig/001.mgz t1-wm-mask-fs.mgz t1-wm-mask-ns.nii 4. Finally, I want to apply the t1 white matter mask to the t2 image using the lta file generated in step 1. for this I use: mri_convert --apply_transform t1-to-t2.lta t1-wm-mask-ns.nii t1-to-t2.nii step 1-3 work fine. but when i try to apply the transform in the last step, the output is zeros. The output of mri_robust_register in step 1 is fine. Does mri_robust_register use mri_convert to apply the transform and create the output image ? Or am I making some error ? thanks Mehul ___ 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.
Re: [Freesurfer] Setting lesion voxels
Hi Ruthger, I use the following command to take the binary lesion mask (native space) to the freesurfer space: mri_convert -rl mri/brainmask.mgz -rt nearest lesion_mask_native_space.mgz lesion_mask_fs_space.mgz I always overlay lesion_mask_fs_space.mgz on brainmask.mgz to check that things look correct. Hope this helps. Mehul On Mon, Oct 11, 2010 at 1:51 AM, Righart, Ruthger Dr. ruthger.righ...@med.uni-muenchen.de wrote: Dear Freesurfers, I would like that surface analyses (i.e., cortical thickness) are not confounded by subcortical lesions in patients. Freesurfer provides a work-around in its release notes ( http://surfer.nmr.mgh.harvard.edu/fswiki/ReleaseNotes), which is to set lesion voxels in brainmask.mgz and norm.mgz to 110 and in aseg.mgz to 77 and then run recon-all -autorecon2-cp (see copy text underneath). As I have lesion masks of every patient I thought it would be best to automatize this step. In that case, I would like to bring mask images with white matter lesions of these patients in the same (freesurfer) space as my T1.mgz, with dimensions 256 x 256 x 256, spacing 1 x 1 x 1., and after that set the lesions to the aforementioned values. Is there a possibility to bring lesion mask of these patients in the same space as T1.mgz in freesurfer? I inspected the function bbregister and mri_convert but could not find a solution here. Any help or suggestions are appreciated! Ruthger Righart (PhD) Institute for Stroke and Dementia Research (ISD) Klinikum der Universität München Max-Lebsche-Platz 30 81377 Munich | Germany ruthger.righ...@med.uni-muenchen.de http://surfer.nmr.mgh.harvard.edu/fswiki/ReleaseNotes • mris_make_surfaces: where wm lesions exist, even if the lesion is filled in wm.mgz volume, when mris_make_surfaces is run to create the final surfaces, because it uses brain.finalsurfs.mgz for its intensity info, and doesnt consider the fill data in wm.mgz, its possible for surfaces to not follow grey matter on the perimeter of the lesion. A fix will appear in a future release. A potential work-around is to set lesion voxels in brainmask.mgz and norm.mgz to 110 and lesions voxels in aseg.mgz to 77 (lesion mask can be derived manually or from an automated algorithm), then run recon-all -autorecon2-cp ___ 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.
Re: [Freesurfer] RAM question
Hi Folks, As Nick suggested 4GB per subject is best. Here is my recent experience when memory resources may be limited and 4gb per subject is not available. I ran FS v5.0 on a cluster on which I was assigned a node with 16 cores and 32gb . At a time, I am charged for all 16 cores on the node, whether I use them all or not; so i am trying to maximize the number of jobs I can submit. I have the option of submitting 8,12,15 or 16 jobs (subjects) on the node (using the -pe flag in the sun grid engine) Depending on the number of jobs the 32gb is divided equally among the jobs; thus, if I submit 16 jobs, each job is assigned 2gb; if I submit 8 jobs, each jobs is assigned 4gb. I found that if I submit 16 or 15 jobs, about half of them fail for me; If I submit 12 jobs, 2.66gb (32/12) is assigned to each subject and in this case most of jobs complete successfully; Out of more than 500 subjects only 3-4 cases failed if I assigned 2.66gb per subject. hope this helps... Mehul On Tue, Sep 21, 2010 at 7:32 AM, Nick Schmansky ni...@nmr.mgh.harvard.eduwrote: the visualization tools will work with just 1GB (or less) ram. its the recon-all stream that consumes a lot of memory. 2GB might work on when processing some subjects, but 4GB is best. it will fail midway through the processing stream if there is not enough memory (during mri_ca_register). n. On Tue, 2010-09-21 at 09:29 -0400, O'Shea, Colin (NIH/NIMH) [V] wrote: I realize that FreeSurfer requires at least 4 GB RAM, and I’m in the process of getting extra ram for my computer. If I install FreeSurfer now, will the program simply not work or just run slow (or something in between)? Thanks, Colin ___ 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 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] setting segmentation brush information from command line in tkmedit
Hi Folks, Currently I set the segmentation brush information from tkmedit buttons. (For example: Set Segmentation as Use as source and also set the color) I need to do this for a large number of subjects and I was wondering if it was possible to do this from command line ? thanks Mehul ___ 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] using -nuintensitycor-3T flag with v-5.0.0
Hi Folks, I would like to use the -nuintensitycor-3T flag as listed in the release notes (http://surfer.nmr.mgh.harvard.edu/fswiki/ReleaseNotes) I had two questions about this new option: 1. Can I just use this as follows: recon-all -s bert -autorecon1 -autorecon2 -nuintensitycor-3T (then run -autorecon3) ? 2. I don't think there are any parameters associated with this flag. is this correct ? thanks Mehul ___ 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] question about the -zgez flag with mri_convert
Hi Folks I use mri_convert to create nifti files from the dicom images from a 3T GE scanner. One of the option for mri_convert is -zgez with the following description from the mri_convert help: -zgez, --zero_ge_z_offset set c_s=0 (appropriate for dicom files from GE machines with isocenter scanning) Before the scan we landmark the subject; that is we set a point on the head coil to 0 and then proceed with the scan. Pardon the naive question, but does isocenter scanning refer to this step ? Also I created two nifti files with and without this flag; the only differences in the two headers were with in quartern_z field. without flag quartern_z = -50.061 and with the flag quartern_z = -90. (and one entry each in srow_z, sform, qform and vox2ras were also affected) Our scan has 180 slices and to me quartern_z = -90 seems the right one. is this correct ? Thanks Mehul ___ 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] registering a segmentation from one time-point to another
Hi Folks, I have two MR scans per subject. Lesions (white matter hypo-intensities) have been manually outlined for the first time-point. I would like to create a lesion mask for the 2nd time-point. But instead of starting from scratch, I would like to use the lesion-mask from the 1st time-point as a starting point for the lesion mask for time-point 2. for this task, I think I can use the following steps: Step 1) Register 1st time-point to the 2nd scan (using 2nd timepoint as reference) with flirt Step 2) Use mri_convert to apply the transformation matrix from flirt to the lesion mask and create a lesion mask for time-point 2. Step 3) make corrections to the lesion mask for time-point 2 if required. Would you suggest to do it in this manner or is an alternative approach recommended ? thanks Mehul ps: I thought of using the longitudinal freesurfer pipeline but i dont know how to compute the transformation matrix from time-point 1 to time-point 2 use the pipeline... ___ 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] creating binary masks for any given structure in aseg.mgz or aparc+aseg.mgz
Hi Folks, I would like to create binary masks for any given structure in aseg.mgz and aparc+aseg.mgz. I would then also like to map the masks back to native space. I think, I should use mri_binarize followed by mri_convert as shown in the two examples below: Example 1: a. mri_binarize --i mri/aseg.mgz --match 10 --o deep-gm-rois/lh-thalamus.mgz b. mri_convert -rl mri/orig/001.mgz -rt nearest deep-gm-rois/lh-thalamus.mgz deep-gm-rois/lh-thalamus.nii Example 2: a. mri_binarize --i mri/aparc+aseg.mgz --match 1022 --o ctx-gm-rois/lh-postcentral.mgz b. mri_convert -rl mri/orig/001.mgz -rt nearest ctx-gm-rois/lh-postcentral.mgz ctx-gm-rois/lh-postcentral.nii Ques 1: Is the best approach for this task ? I noticed in the FAQ that mri_vol2vol is recommended for conversion from freesurfer space to anantomical space. Ques 2: Is it better to use mri_vol2vol than mri_convert with the reslice option ? Thanks Mehul ___ 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.
Re: [Freesurfer] creating binary masks for any given structure in aseg.mgz or aparc+aseg.mgz
Thanks Doug. Using mri_label2vol is a good idea and I will use this approach. Now to create the individual binary masks I just need to use mri_label2vol, followed by mri_binarize: For example: a. mri_label2vol --seg aparc+aseg.mgz --regheader rawavg.mgz --o aparc+aseg-in-rawavg.mgz --temp aparc+aseg.mgz b. mri_binarize --i aparc+aseg-in-rawavg.mgz --match 1022 --o lh-postcentral.nii Also, when I looked at the help, I first thought I could do the same thing with mri_annotation2label followed by mri_label2vol . for example: a. mri_annotation2label --subject bert --hemi lh --annotation aparc --outdir labels b. mri_label2vol --label ../labels/lh.postcentral.label --regheader rawavg.mgz --o lh-postcentral.nii --temp aparc+aseg.mgz But when I checks the results, (from using mri_annotation2label followed by mri_label2vol), the individual binary mask looks incorrect Did I make an error in the second approach ? Thanks Mehul On Tue, Jul 13, 2010 at 2:41 PM, Douglas N Greve gr...@nmr.mgh.harvard.eduwrote: I think either should work, though I am more comfortable with mri_vol2vol. Depending on what you are doing you might want to use mri_label2vol. This can convert the aseg.mgz directly to the native anatomical space. The main difference is that it will resolve boundary voxels by voting. With your method, you will have some voxels that will be in more than one label, which may or may not be a problem depending on what you are doing. doug Mehul Sampat wrote: Hi Folks, I would like to create binary masks for any given structure in aseg.mgz and aparc+aseg.mgz. I would then also like to map the masks back to native space. I think, I should use mri_binarize followed by mri_convert as shown in the two examples below: Example 1: a. mri_binarize --i mri/aseg.mgz --match 10 --o deep-gm-rois/lh-thalamus.mgz b. mri_convert -rl mri/orig/001.mgz -rt nearest deep-gm-rois/lh-thalamus.mgz deep-gm-rois/lh-thalamus.nii Example 2: a. mri_binarize --i mri/aparc+aseg.mgz --match 1022 --o ctx-gm-rois/lh-postcentral.mgz b. mri_convert -rl mri/orig/001.mgz -rt nearest ctx-gm-rois/lh-postcentral.mgz ctx-gm-rois/lh-postcentral.nii Ques 1: Is the best approach for this task ? I noticed in the FAQ that mri_vol2vol is recommended for conversion from freesurfer space to anantomical space. Ques 2: Is it better to use mri_vol2vol than mri_convert with the reslice option ? Thanks Mehul ___ 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 gr...@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 ___ 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] probabilistic white matter mask
Hi Folks, I believe that I can create a binary white matter mask using mri_wmfilter. I was wondering if there is a way to create a probabilistic white matter mask ? (for each voxel, I would like to obtain the probability of it being white matter) I did a search on the mailing list and found the following related message: http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg11786.html The message suggests to use -write_probs flag with the mri_ca_label command but I dont see this flag as on option in the usage instructions for mri_ca_label. If this is the right way, should i just add -write_probs wm.mgz to mri_ca_label to get a probablistic white matter mask ? Thanks Mehul ___ 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] Changing tkmedit settings
Hi Folks, In tkmedit, the keyboard commands to view main and aux volume are 'Ctrl+1' and Ctrl+2'. Is it possible for me to change them to just '1' and '2' respectively ? (like how only 'n' is required for the navigation tool) Thanks Mehul ___ 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] areas to be excluded during manual corrections of pial and main surfaces
Hi Folks, On the FAQ, the response to one question says that the surfaces near the medial wall, hippocampus and amygdala are unreliable and should be excluded from a thickness analysis. The link for this is: http://surfer.nmr.mgh.harvard.edu/fswiki/UserContributions/FAQ#Q.Thesurfacesnearthemedialwall.2Chippocampus.2Candamygdalaaren.27taccuratelyfollowingthestructuresthere.HowcanIfixthis.3F My question is, as a rule of thumb, are there any other regions that should be excluded while doing manual corrections to pial and main surfaces ? For example, should we stop looking at the pial and main surfaces in the axial slices where the medulla is observed ? Basically, are there any other suggestions about which regions do not need corrections to the pial and main surfaces ? thanks Mehul ___ 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] passing a WM lesion mask to recon-all
Hi Folks, We have some subjects with very large white matter (WM) lesion loads. We noticed that the pial and main are pial surfaces are not affected by WM lesions. However, they seem to be affection by the juxta-cortical WM lesions (which have dark intensity) In the past, my colleagues have manually outlined the lesions and created a binary lesion masks. I was wondering if : 1. I can pass this white matter lesion mask as input to recon-all as an input variable ? essentially the idea is to find a way to make recon-all to consider these as white matter (i am only interested in the cortical thickness measurements and not the WM volume so i think it is okay to do so ) 2. if this is not possible, can i use this white matter lesion mask to edit wm.mgz ? in this option, i could find the locations of the non-zero pixels of the lesion mask and then set the corresponding locations in wm.mgz to 110. Does this sound like a bad idea ? or would you suggest a different way to do this ? thanks Mehul ___ 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] creating a binary mask volume
Hi Folks, I would like to create a binary mask volume for the Right-Thalamus (all voxels of right-thalamus as 1 and 0 otherwise). I looked at this help page http://surfer.nmr.mgh.harvard.edu/fswiki/mri_vol2roi and i think I should use the following command described on this page: mri_vol2roi --label your.label --srcvol f --srcreg register.dat --finalmskvol labelbinmask --roiavg /tmp/not.wanted.dat is this correct ? also could someone give me an example? i am not sure what all of the parameters are. thanks Mehul ___ 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.
Re: [Freesurfer] creating a binary mask volume
Thanks Bruce. Following that, to convert the rth.mgz back to native space I think I can use the following command: mri_convert -rl orig/001.mgz -rt nearest rth.mgz rth.nii thanks Mehul On Sun, Apr 11, 2010 at 1:38 PM, Bruce Fischl fis...@nmr.mgh.harvard.eduwrote: Hi Mehul, try mri_binarize instead. It would be something like: mri_binarize --match 49 --i aseg.mgz --o rth.mgz where 49 is the index for right thalamus proper (from the FreeSurferColorLUT.txt file) and rth.mgz is the output volume cheers, Bruce On Sun, 11 Apr 2010, Mehul Sampat wrote: Hi Folks, I would like to create a binary mask volume for the Right-Thalamus (all voxels of right-thalamus as 1 and 0 otherwise). I looked at this help page http://surfer.nmr.mgh.harvard.edu/fswiki/mri_vol2roi and i think I should use the following command described on this page: mri_vol2roi --label your.label --srcvol f --srcreg register.dat --finalmskvol labelbinmask --roiavg /tmp/not.wanted.dat is this correct ? also could someone give me an example? i am not sure what all of the parameters are. thanks Mehul 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] generating WM parcellation using Destrieux atlas
Hi Folks, I was trying to help a colleague generate a wmparc for the Destrieux atlas. That is I am trying to create wmparc.a2009s.mgz and wmparc.a2009s.stats I used the following two commands: mri_aparc2aseg --s subject_id --labelwm --hypo-as-wm --rip-unknown --new-ribbon --wmparc-dmax 4 --o mri/wmparc.a2009s.mgz --ctxseg aparc.a2009s+aseg.mgz mri_segstats --seg mri/wmparc.a2009s.mgz --sum stats/wmparc.a2009s.stats --pv mri/norm.mgz --excludeid 0 --brain-vol-from-seg --brainmask mri/brainmask.mgz --in mri/norm.mgz --in-intensity-name norm --in-intensity-units MR --etiv --subject subject_id --surf-wm-vol --ctab $FREESURFER_HOME/FreeSurferColorLUT.txt I was able to run this and I noticed that a number of labels in wmparc.a2009s.stats do not have any voxels assigned to them. for examples: 762 4174 00.0 wm-rh-S_subcentral_ant 0. 0. 0. 0. 0. 763 4175 00.0 wm-rh-S_subcentral_post 0. 0. 0. 0. 0. 764 4176 00.0 wm-rh-S_suborbital 0. 0. 0. 0. 0. 765 4177 00.0 wm-rh-S_subparietal 0. 0. 0. 0. 0. 766 4178 00.0 wm-rh-S_supracingulate 0. 0. 0. 0. 0. 768 4180 00.0 wm-rh-S_temporal_superior0. 0. 0. 0. 0. 769 4181 00.0 wm-rh-S_temporal_transverse0. 0. 0. 0. 0. It seems like only labels corresponding to the Desikan atlas have voxels assigned to them in wmparc.a2009s.stats ? do you think I made an error while generating wmparc.a2009s.stats ? thanks Mehul ___ 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.
Re: [Freesurfer] modification made to longitudinal results
Hi Martin, Just wanted to clarify one question about edits to the longitudinal pipeline. Lets say there is a data-set with 200 subjects with 5 time-points each (1000 scans total). I create the base image and i would only like to make corrections in the 200 base images (if required) and not in the cross-sectional time-points. My understanding from an earlier thread was that this should be fine. But in this thread you said control points are copied from the 'cross-sectional by default' So is the following correct? I should only make edits to the wm.mgz (from the base) and avoid adding any control points to the base ? I apologize for the repeated questions on the issue, but just wanted to make sure that I was making the corrections correctly. Thanks Mehul On Thu, Mar 11, 2010 at 7:20 PM, Martin Reuter mreu...@nmr.mgh.harvard.eduwrote: Derin OK, Allison discovered it in the depth of recon-all :-) The control points are copied from the cross sectional by default (and should be edited there). There is an optional flag to get them from the base, but that has never been really tested. It only makes sense if there is very little structural change across time. But might save people some editing time, if they have 10 images with almost no WM change. We'll probably do a similar setup for the WM edits in the next release. Cheers, Martin On Wed, 2010-03-10 at 15:16 -0600, Derin Cobia wrote: Martin, Not to add more to the confusion, but what if the final longitudinal scans are not fine and need some editing. Can they be rerun in the standard way to incorporate edits (e.g., -autorecon2-cp -autorecon3 ... ), or should something different be done? Will rerunning these edited long scans in the standard fashion disrupt them in some way? My hunch is that it won't, but I want to make sure. Thanks. -Derin On Mar 10, 2010, at 2:17 PM, Martin Reuter wrote: OK I get it. There was a confusion about adding control points and adding new time points. My answer concerns adding new time points, so you can ignore it. Do not add any edited results as new time points. Concerning your question see the discussion on this list from Feb 12 that treats exactly this topic. In short: - edit the cross sectionals (you have done that) - run the base and edit the base - then the longitudinals should be fine Martin On Wed, 2010-03-10 at 13:08 -0600, Guang Zeng wrote: Hello, Martin, Thanks a lot for your reply, but I am still not very clear about few issues. I think that is because of my unclear description. What happen is: I have two scans which I want to do longitudinal analysis, however, when I finish the cross-sectional analysis, I found the results are not so good because of the low contrast between white matter and gray matter. I added control points to these two scans, rerun them. The results looks much better now, then I go to the longitudinal stream. However, when I load the longitudinal results, I found the kind of problem happens again (lots of no-label region in superior frontal). So I added control points to the longitudinal results directly, and I want rerun them. Based on your reply, I need consider those longitudinal results which I added control points to as new timepoints, rerun them cross-sectionally again, is it correct? Thanks! Guang Here, I just want to add some control points to the FreeSurfer longitudinal results, not new time points. Subject: Re: [Freesurfer] modification made to longitudinal results From: mreu...@nmr.mgh.harvard.edu To: freesurfer...@hotmail.com CC: freesurfer@nmr.mgh.harvard.edu Date: Thu, 25 Feb 2010 10:33:36 -0500 Hi Guang, Depending on what you do you can choose different routes. Note, for both of these you first need to run the new timepoint cross sectionally (step 1 in the description http://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalProcessing ): recon-all -all -s newtpid -i path/to/dicom Here are the two options once the cross sectional results are there: 1. you have only very few timepoints in the base/template so far (2 or 3). In those cases I would recommend to rerun the base and rerun all the longitudinals with the new and more accurate base. The commands are on the Wiki and are the same as usual. 2. you have many time points in your base, the additional time point is not likely to change the base much. In that case you can simply 'patch' the base without reprocessing and only run a single longitudinal run. Let me know if you want to go route 2 because I am writing a script to patch the base (there are a few files that need to be added so that the longitudinal run will go through). If there is demand, I will put priority on this and make it available. Best, Martin On Thu,
[Freesurfer] using control points versus editing wm.mgz
Hi Folks, I understand that control points should be used to fix errors due to bad skull stripping or intensity normalization (are there any other scenarios when one should use control points ?) I followed the instructions on making corrections for issues due to intensity normalization. http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/ControlPoints As I understand, one can fix issues due to intensity normalization either by using control points or editing wm.mgz. I would like to only edit wm.mgz and not use control points. Is this okay ? In your experience, are there situations, where this would be a bad idea ? Thanks Mehul ps: The motivation for this is that I want to use the longitudinal processing stream and I only want to edit the base and not edit the cross-sectional time-points. ___ 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.
Re: [Freesurfer] Longitudinal processing clarification
Hi Nick (and Martin) Could you please clarify if the edits come only from the base or if they are copied from the cross-sectionals ? It would be great if I only had to make edit in the base . We have 5 timepoints for each subject (each a year apart) and i want to use the results from the longitudinal pipeline only. If I did not need to edit the cross-sectional timepoints and if editing only the base was okay this would be very helpful. Martin, do you normally only make edits to base when using the longitudinal pipeline ? Thanks Mehul On Fri, Feb 12, 2010 at 6:01 PM, Martin Reuter mreu...@nmr.mgh.harvard.eduwrote: Hi Derin, this is how I see it: The base might need edits and if it does, they should be made. As you say, the surfaces of the longitudinal runs are initialized with the surface of the base, so, if it is not correct, this will pollute all time points. Another reason to edit the base is that wm edits (255) and deletes (1) are copied from the base to all time points, as you mentioned. Usually this should be sufficient, but maybe you need further edits in a specific longitudinal run, never happened to me. The old longitudinal processing stream really did not have any cross sectional runs. Instead of the base they processed TP1 (introducing a bias) and copied the edits from there to the longitudinals. We now copy it from the base. Hope that helps. Nick mentioned that edits are copied from the cross sectionals, but I think they only come from the base. At least that is what I think was done and therefore put it on the wiki. Maybe Nick can verify what really goes on in recon-all as he programmed that part? It might make sense to edit the cross sectionals anyway because: - you then get more reliable cross sectional results (to compare to etc). - the cross sectional asegs are 'fused' (probabilistic voting), to initialize the labeling in the longitudinal runs, so more accuracy in the cross sectionals cannot hurt. However I never really found this to be necessary, but my data was usually high quality. Best, Martin On Fri, 2010-02-12 at 13:51 -0600, Derin Cobia wrote: Bumping this back up to the list. Could someone please address? To be more concise, during longitudinal processing, where are wm/gm edits made? On the cross-sectional runs, to the base, or on the .long. runs themselves? Thanks. -Derin Begin forwarded message: From: Derin Cobia d-co...@northwestern.edu Date: February 1, 2010 3:54:46 PM CST To: freesurfer@nmr.mgh.harvard.edu Help freesurfer@nmr.mgh.harvard.edu Subject: Longitudinal processing clarification (Questions for all, but maybe more directed towards Martin) I have some questions about manual edits in regard to the longitudinal stream. Based on the current info on the wiki page, it appears that no edits are pushed forward from the cross-sectional (initial) runs of the subject data, as any information on the *.long.* runs come from the template. However, I've noticed that WM edits and surfs from the template are used for the *.long.* processing. In essence, my question is when/where should manual edits (i.e., control points, WM, pial edits) occur when conducting longitudinal processing in 4.5? It seems I remember that in 4.3 edits in the cross-sectional work would be pushed forward to the *.long.* images, but such is not the case anymore (?). Are edits to occur in the template (i.e., 'base') now, or in the *.long.* runs? Thanks. -Derin ___ 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] recon-all -qcache question
Hi Guang, It is much faster as it does not go through the whole autorecon1, 2, 3 process. For us, it take about 45 minutes per subject. Regards Mehul On Mon, Nov 30, 2009 at 8:04 PM, Guang Zeng freesurfer...@hotmail.comwrote: Hi, there, I have 60 subjects which have been FS analyzed without the flag -qcache. Now, I wan to do group analysis using QDEC. I need run recon-all -s subjectid -qcache for each of my subjects. Just want to know how it will take to finish the recon-all -s subjectid -qcache command for one subject? Will it go through the whole -autorecon1, 2, 3 process again? Thanks a lot! Guang -- Hotmail: Trusted email with powerful SPAM protection. Sign up now.http://clk.atdmt.com/GBL/go/177141665/direct/01/ ___ 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] script to check multiple cases with tksurfer
Hi Folks, I would like to visually check the cortical parcellation for a large number of patients (200) using tksurfer. Is there a script to which I could submit a list of subject id's and get it to launch tksurfer and display the parcellation data for each subject sequentially ? Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
[Freesurfer] white matter mask
Hi Folks, I want to create a white matter mask (a volume in which all of the white matter voxels are 1 and 0 otherwise) For this, I am planning to use wmparc.mgz and then set the voxels with labels of wm-lh-? and wm-rh-? as 1. (here wm-lh-? represents labels such wm-lh-temporalpole, wh-lh-insula etc) From wmparc.stats I found there are 35 such labels for each hemisphere. Is this the right way to create the white matter mask ? Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] white matter mask
I just realized that a simpler way to create this mask would be to use ribbon.mgz and find all the voxels with the labels 2 and 41 which correspond to Left_White and Right_White. https://surfer.nmr.mgh.harvard.edu/fswiki/mris%25volmask Does this sound correct ? Mehul On Sun, Nov 8, 2009 at 1:48 PM, Mehul Sampat mpsam...@gmail.com wrote: Hi Folks, I want to create a white matter mask (a volume in which all of the white matter voxels are 1 and 0 otherwise) For this, I am planning to use wmparc.mgz and then set the voxels with labels of wm-lh-? and wm-rh-? as 1. (here wm-lh-? represents labels such wm-lh-temporalpole, wh-lh-insula etc) From wmparc.stats I found there are 35 such labels for each hemisphere. Is this the right way to create the white matter mask ? Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
[Freesurfer] reslicing question
Hi Folks In my dataset each MRI exam is 256*256*180. I would like to convert wmparc.mgz back to 256*256*180. Since this is the segmentation output, I think I should use set the resample type to nearest and use a command like mri_convert -rl mri/orig/001.mgz -rt nearest mri/wmparc.mgz mri/wmparc.nii Does this look correct or is there anything else I need to specify? Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] comparison of mean thickness of gyri in two groups
Thank you for your suggestions Bruce and Mike. Mike, could you please provide the motivation behind your suggestion of using mean cortical thickness as a covariate ? Let's consider this is the context of a simple example. Let's say we compare the mean thickness of the pre-central gyrus in patients versus controls. Does the biology suggest that even after one accounts for age and gender, there are natural variations in the gyrus thickness measurements ( that is, some people would just happen to have larger thickness measurements ) ? If this is the case, then one could normalize the thickness of the individual gyrus measurement (pre-central gyrus in this example) with the mean cortical thickness. However, in most of the papers I have only seen age and gender used as co-variates. Thanks Mehul On Sun, Nov 1, 2009 at 4:44 PM, Michael Harms mha...@conte.wustl.eduwrote: To me, it makes much more sense to use mean cortical thickness as a covariate for thickness-based analyses. cheers, Mike H. Hi Mehul, the MNI group had an abstract showing that thickness didn't need eTIV correction at HBM a number of years ago, and it has been our experience as well. cheers, Bruce On Sat, 31 Oct 2009, Mehul Sampat wrote: Hi Folks, If am comparing the mean thickness for certain gyri (pre-central, post-central) in a controls versus patients. I obtain the thickness measurements from lh.aparc.a2009s.stats and rh.aparc.a2009s.stats I was wondering if I need to normalize these thickness measurement with the estimated total intracranial volume (eTIV) ? Or has anyone reported that the these are independent of eTIV ? If so then as I understand I would only need to account for age and gender in any subsequent analysis using GLM. Thanks Mehul ___ 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] comparison of mean thickness of gyri in two groups
Hi Folks, If am comparing the mean thickness for certain gyri (pre-central, post-central) in a controls versus patients. I obtain the thickness measurements from lh.aparc.a2009s.stats and rh.aparc.a2009s.stats I was wondering if I need to normalize these thickness measurement with the estimated total intracranial volume (eTIV) ? Or has anyone reported that the these are independent of eTIV ? If so then as I understand I would only need to account for age and gender in any subsequent analysis using GLM. Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
[Freesurfer] question about load_segstats
Hi FS folks I was trying to use the load_segstats function to read aseg.stats and wmparc.stats It works well will these files but I get the following error if I try to read lh.aparc.stats or rh.aparc.stats [segname segindex segstats] = load_segstats('lh.aparc.stats','bert'); ??? Subscripted assignment dimension mismatch. Error in == load_segstats at 113 segstats(nthrow,:) = [nvox vol segmn segstd segmin segmax segrng]; Any suggestions on how I could fix the matlab function ? Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] GLM question
Thanks Doug, I have two follow-up questions. 1. I am using QDEC and my discrete variables are: (a) TYPE (PATIENTS/CONTROLS) and (b) GENDER I include age as a continous co-variate and then select all of the variable before running the design. I believe QDEC is generating the various possible contrast matrices and the corresponding research question. When I run QDEC, I see the following questions: A. Does the average thickness, accounting for Gender differ between Patients and Controls? B. Does the thickness-age correlation, accounting for Gender differ between Patients and Controls ? (I think this is looking at whether the slopes are different for the two groups) I was wondering why I dont see a question like Does the average thickness, accounting for Gender *AND AGE* differ between Patients and Controls? 2. It is likely I dont have a full understanding of the GLM theory. Could you please suggest some good references describing the GLM theory ? Thanks Mehul On Thu, Oct 29, 2009 at 10:57 AM, Douglas N Greve gr...@nmr.mgh.harvard.edu wrote: Mehul Sampat wrote: Hi FS folks, I have a basic GLM question. I went through the tutorials online but I was not sure and wanted to check with someone. I am trying to compare the cortical thickness between a group of patients (n = 166) and controls (n = 76). For patients mean age is 49.8 +/- 9.1 and there are 55 Male; 111 Female For controls mean age is 40.5 +/- 11.4 and there are 26 Male; 50 Female If include gender as a fixed factor does the output of the GLM answer the following: 1. Is the cortical thickness different between the patients and controls accounting for gender All the factors in mri_glmfit/QDEC are random factors. But, yes, it would answer that question. 2. As I understand the GLM setup one can control for gender and other discrete factors but not for continuous co-variates such as age ? That is one can find the association between thickness and age and see if it is different for the two groups. However if the age distributions for the two groups are different one cannot control for it with GLM. is this interpretation correct ? Not quite. You can always put age in as a continuous covariate. If there are effects of age or an interaction between age and group, then there are some subtle statistical issues. doug If so, how would one control for age in such an analysis ? Any help is much appreciated. Thanks Mehul 2. For the correction of multiple comparisons, when should one use FDR as compared to monte-carlo simulations ? Thanks Mehul ___ 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 gr...@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422 In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
[Freesurfer] GLM question
Hi FS folks, I have a basic GLM question. I went through the tutorials online but I was not sure and wanted to check with someone. I am trying to compare the cortical thickness between a group of patients (n = 166) and controls (n = 76). For patients mean age is 49.8 +/- 9.1 and there are 55 Male; 111 Female For controls mean age is 40.5 +/- 11.4 and there are 26 Male; 50 Female If include gender as a fixed factor does the output of the GLM answer the following: 1. Is the cortical thickness different between the patients and controls accounting for gender 2. As I understand the GLM setup one can control for gender and other discrete factors but not for continuous co-variates such as age ? That is one can find the association between thickness and age and see if it is different for the two groups. However if the age distributions for the two groups are different one cannot control for it with GLM. is this interpretation correct ? If so, how would one control for age in such an analysis ? Any help is much appreciated. Thanks Mehul 2. For the correction of multiple comparisons, when should one use FDR as compared to monte-carlo simulations ? Thanks Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
[Freesurfer] multi-channel lesion segmentation.
Hi all, I would like to segment lesions using both T1 and FLAIR images. I think this is not possible with the current version of freesurfer. Is that right ? Could I somehow combine the segmentation results from aseg with the FLAIR images ? Regards Mehul ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Re: [Freesurfer] Segmentation
Hi I am interested in measuring the Brain Parenchymal Fraction (BPF) which is defined as follows: BPF = (total White Matter (WM) Volume + total Gray Matter (GM) Volume) / (total White Matter Volume (WM) + total Gray Matter Volume (GM)+ total CSF volume) I think I get the White and Gray Matter Volumes for each hemisphere from the command: mris_anatomical_stats. I think the denominator in the BPF equation is just the Intracranial Volume, reported in the aseg.stats file Would this be an appropriate way to measure these volumes ? Thanks Mehul p.s: For the subject bert the BPF = 1661478/1799364 = 0.92 On 6/25/07, Bruce Fischl [EMAIL PROTECTED] wrote: Hi Antonio, that's why we we use the surfaces for white matter and cortical gray matter volume. cheers, Bruce On Mon, 25 Jun 2007, Gallo, Antonio (NIH/NINDS) [F] wrote: Hi Bruce, Thank you for your quick response. Our problem, actually, is that when we visualize the aseg.mgz volume (with color-coded subcortical structures) we see that WM is classified as GM (left circle around insula in the figure) and GM is classified as WM (right lower circle in the figure). So, even if the surfaces look good, we are worried about the effects of this misclassification on volume measurements in aseg.mgz. Thanks, Antonio -Original Message- From: Bruce Fischl [mailto:[EMAIL PROTECTED] Sent: Monday, June 25, 2007 4:10 PM To: Gallo, Antonio (NIH/NINDS) [F] Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Segmentation Hi Antonio, I wouldn't use the wm.mgz for much of anything. It's really just an intermediate step in the creation of the surfaces. In general we trust the ?h.white and ?h.pial surfaces first, then the aseg.mgz for things like hippocampus, ventricles, etc... Bruce On Mon, 25 Jun 2007, Gallo, Antonio (NIH/NINDS) [F] wrote: Hi All, Upon reviewing subortical segmentations I see that often a mismatch exists between what appears on the wm.mgz and the color-coded segmented maps. Precisely - as an example - in the attached figure you can see that in the case of left insula WM, part of the tissue was classified as GM (in brown) in the color-coded segmented map and as WM in the wm.mgz. Contrariwise, the right temporal-parietal GM cortex part of the tissue is classified as WM (in green) the color-coded segmented map and as non-WM in the wm.mgz. The latter occurs in many regions of the brain without a clear anatomical pattern. I was wondering if this is just a display issue or it rather reflects some mismatch in actual tissue classification with consequential errors in the generated values of thickness and volumes as well? If so, is there any way I can correct for this? Thank you in advance, Antonio Antonio Gallo, MD NIB-NINDS-NIH 10 Center Drive Building 10, Room 5B16 Bethesda, MD, 20892 - USA ph #: 001-301-402.6391 fax #: 001-301-402.0373 *** ___ 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] Segmentation
Hi Bruce, Thanks for the quick response. Are there plans to add functionality to label sulcal CSF from T2 images? (Or can I do this easily using some existing script ?) My intuition is that this would be helpful since as brain tissue atrophy occurs in a given neurological disorder (Multiple Sclerosis, aging), one would expect both the ventricular and the sulcal CSF volumes to increase and that we would get a lower estimate of atrophy by measuring only the ventricular CSF ? Regards, Mehul On 7/19/07, Bruce Fischl [EMAIL PROTECTED] wrote: Hi Mehul, We explicitly label ventricular CSF, but not sulcal CSF (which you pretty much can't do from only a T1 volume). The ICV would give you an approximation of the sum of it all, so yes I think what you are doing is a reasonable approach. cheers, Bruce On Thu, 19 Jul 2007, Mehul Sampat wrote: Hi I am interested in measuring the Brain Parenchymal Fraction (BPF) which is defined as follows: BPF = (total White Matter (WM) Volume + total Gray Matter (GM) Volume) / (total White Matter Volume (WM) + total Gray Matter Volume (GM)+ total CSF volume) I think I get the White and Gray Matter Volumes for each hemisphere from the command: mris_anatomical_stats. I think the denominator in the BPF equation is just the Intracranial Volume, reported in the aseg.stats file Would this be an appropriate way to measure these volumes ? Thanks Mehul p.s: For the subject bert the BPF = 1661478/1799364 = 0.92 On 6/25/07, Bruce Fischl [EMAIL PROTECTED] wrote: Hi Antonio, that's why we we use the surfaces for white matter and cortical gray matter volume. cheers, Bruce On Mon, 25 Jun 2007, Gallo, Antonio (NIH/NINDS) [F] wrote: Hi Bruce, Thank you for your quick response. Our problem, actually, is that when we visualize the aseg.mgz volume (with color-coded subcortical structures) we see that WM is classified as GM (left circle around insula in the figure) and GM is classified as WM (right lower circle in the figure). So, even if the surfaces look good, we are worried about the effects of this misclassification on volume measurements in aseg.mgz. Thanks, Antonio -Original Message- From: Bruce Fischl [mailto:[EMAIL PROTECTED] ] Sent: Monday, June 25, 2007 4:10 PM To: Gallo, Antonio (NIH/NINDS) [F] Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Segmentation Hi Antonio, I wouldn't use the wm.mgz for much of anything. It's really just an intermediate step in the creation of the surfaces. In general we trust the ?h.white and ?h.pial surfaces first, then the aseg.mgz for things like hippocampus, ventricles, etc... Bruce On Mon, 25 Jun 2007, Gallo, Antonio (NIH/NINDS) [F] wrote: Hi All, Upon reviewing subortical segmentations I see that often a mismatch exists between what appears on the wm.mgz and the color-coded segmented maps. Precisely - as an example - in the attached figure you can see that in the case of left insula WM, part of the tissue was classified as GM (in brown) in the color-coded segmented map and as WM in the wm.mgz. Contrariwise, the right temporal-parietal GM cortex part of the tissue is classified as WM (in green) the color-coded segmented map and as non-WM in the wm.mgz. The latter occurs in many regions of the brain without a clear anatomical pattern. I was wondering if this is just a display issue or it rather reflects some mismatch in actual tissue classification with consequential errors in the generated values of thickness and volumes as well? If so, is there any way I can correct for this? Thank you in advance, Antonio Antonio Gallo, MD NIB-NINDS-NIH 10 Center Drive Building 10, Room 5B16 Bethesda, MD, 20892 - USA ph #: 001-301-402.6391 fax #: 001-301-402.0373 *** ___ 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 ___ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer