Freesurfer

freesurfer@nmr.mgh.harvard.edu

5 participants
30076 discussions

Surfaces, asegs, aparcs, Subcortical Structures
by James N. Porter 10 Dec '07

10 Dec '07

Hello FreeSurfer Folks- I've seen several threads in the archives discussing the expected inaccuracies in the pial, white, and orig surfaces running through subcortical structures. However, I've been unable to ascertain exactly how much and what kind of inaccuracy is acceptable. It makes sense to me that if the entirety of a structure, say the amygdala, were included or excluded in the pial surface, then it would be easy to adjust later measurements. However, what we see in all of our subjects is a random pattern of half in/half out, with weird shapes, when it comes to the surfaces running in and around subcortical structures. For example, take a look the image at this link. http://www.tc.umn.edu/~norb0062/Amygdala.tiff There you will see the how the hippocampus and amygdala are both half included in the pial surface. Item 1 has a strange guitar-shaped chunk taken out of the amygdala, and item 2 has a triangle and a peanut-shaped area removed. Are these the types of expected and accepted inaccuracies, or is there something afoul here that needs to be corrected? I have one other question regarding the interactions of aseg and surface information. Item 3 in the image points out a section of apparent subdural matter that has been classified as grey matter in the aseg but has been excluded from the pial surface. When the final statistics are spit out of recon-all, is there a section of processing that says something like, "Based on the aseg values my grey matter volume is X, but since Y of that is outside of the pial surface I'll call my total volume X-Y"? Or, is this another item to that requires correction and re-running of recon-all? Many thanks in advance, -- Jim Porter Graduate Student Clinical Science & Psychopathology Research University of Minnesota

4 7

surface problems around insula
by Michael Harms 10 Dec '07

10 Dec '07

Hello, We have observed several instances now where the surfaces around the insula are incorrectly constructed, with the pial surface being pulled inward toward the putamen. The likely source of this problem is that there is no obvious white matter between the putamen and insular cortex. Based on a previous suggestion, we tried editing the aseg to add a thin strip of putative white matter between the labels for the putamen and cortex, but this didn't prevent the problem. jpegs of the brainmask, aseg, and wm volumes are attached to illustrate. These data were processed using FS v4.0.1, EXCEPT that we used the most recent development version of mris_make_surfaces (v 1.105, built 11/2/2007) that we could find at the ftp site. Two questions: 1) Should we be concerned about fixing this type of pial surface problem in the insular cortex? (i.e., since this region typically seems to get labelled as 'unknown' in the aparc parcellation should we just ignore defects of this type?) 2) What controls the degree to which the ?h.white surfaces are allowed to deviate from the ?h.orig surfaces? In particular, while the orig surface is reasonable for this brain in the putamen/insular region (because it follows wm.mgz, which we can edit if need be), the putative white surface is too medial. The notes in 'recon-all' for the 'finalsurfs' stage state that "The white surface is created by "nudging" the orig surface so that it closely follows the white-gray intensity gradient as found in the T1 volume". How does one prevent the white surface from being "nudged" too far away from the orig surface, as occurred for this brain? thanks, Mike Harms -- 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 6615 Tel: 314-747-6173 660 South Euclid Ave. Fax: 314-747-2182 St. Louis, MO 63110 Email: mharms(a)wustl.edu --------------------------------------------------------------------

1 0

ROI labels for geometry vertices
by Sergey Plis 10 Dec '07

10 Dec '07

Hi, Please, help me to find out how to extract information I need. For each vertex in the geometry of the cortical surface I want an additional label that tells me which ROI this vertex belongs to. I extract all geometrical information using: mne_read_surfaces in matlab. This, however, does not give me the labels for ROI of each vertex. I plan to use freesurfer atlas for ROIs, but if anyone has a general answer that is suitable for any atlas I will be happy. A step by step guide would be appreciated :) Thanks! -- Sergey

2 5

wmparc/white matter problems?
by Ryan Muetzel 08 Dec '07

08 Dec '07

Hello, We are running Freesurfer version 4.0.1 and seem to be having problems with the white matter/gray matter parcellation borders. When viewing the pial and the main white surfaces in tkmedit, everything looks good against the t1 brain. However, when the wmparc is loaded on, we are seeing that the wmparc ROIs sometimes do not follow the white surface----the ROIs are going into the gray matter regions. Are the white matter surfaces converted to volumes and then convolved with the aseg/aparcs/etc or are they only used for the cortical thickness computations? It almost seems as if the intensity based wm seg is used and not the wm surface based (?h.dwhite.ribbon.mgz)? Any help is much appreciated! -Ryan

2 4

mri_convert and VB15
by Jason Webster 07 Dec '07

07 Dec '07

Hello, After upgrading our scanner software to VB15A, we began receiving the following error from mri_convert when running unpacksdcmdir: ERROR: incorreclty formatted version string syngo MR B15 found in dicom tag 18,1020 (len = 3 != 6) Numaris Version: syngo MR B15 Maj = -1073750848, Min=-1073771672, MinMin = -1208005856 child process exited abnormally ERROR: mri_convert The error seems to be arising from the parsing of the software version in the DICOM metadata; below are pre and post versions of the relevant line: Software Version : syngo MR B13 4VB13A Software Version : syngo MR B15 >From what I have been told, this is the official version and the metadata from the MGH scanner is the same. Has anyone else encountered this? We are currently using the unpacksdcmdir and mri_convert from v3.0.5 - are there other versions of these that will read DICOMs from VB15? Thanks, Jason

2 1

buglet in fast_inorm
by Sebastian Moeller 07 Dec '07

07 Dec '07

Dear fellow freesurfers, using FreeSurfer 3.0.5 and 4.0.1, I can make fast_inorm.m (as called by inorm) die, with otherwise innocently looking inputs. The syndrom is that if my fist slice contains only one iunder voxel twf_under (see 3.0.5's line 88 of fast_inorm.m) is initialised as scalar (that is not an array), any subsequent try to store a n_timepoint long vector in twf_under just makes matlab choke. This same combination of input dimensionality is fixed for twf_over in the same code. Applying the same special casing of single voxels for iunder solves my problems for good. Here is the meat of the change: if(nthslice,1) > 0) twf_under(nthslice,:) = mean(y2(iunder,:)); else with the following construct: if(nunder(nthslice,1) > 1) twf_under(nthslice, 1) = mean(y2(iunder,:)); elseif (nunder(nthslice,1) == 1) twf_under(nthslice, 1) = y2(iunder,:); else Ahoi & thanks supplying powerful tools Sebastian -- Sebastian Moeller Tel.: 04 21 - 2 18 - 78 38 oder 96 91 Fax.: 04 21 - 2 18 - 90 04 GSM: 0 15 77 - 1 90 31 41 moeller(a)brain.uni-bremen.de AG Kreiter / FB 2 Institut fuer Hirnforschung III Abteilung Theoretische Neurobiologie Universitaet Bremen Biogarten Hochschulring 16a Postfach 33 04 40 28359 Bremen

1 0

retinotopy help
by anisha narula 07 Dec '07

07 Dec '07

Hello, I am carrying a retinotopy analysis and not getting good flattened images. When I view the slices using the sliceview-sess -sf sessid -df sessdir -analysis rtopy -c eccen -map h -slice h command, there is apparent activation the region desired.I then created the maps using the paint-sess command. However when I view the results on the 3D volume using the surf-sess -sf sessid -df sessdir -analysis rtopy -retinotopy eccen command, the retinotopy is messed up. There seems to be some error in the registering of the functionals to the 3D volume. Perhaps I am missing a command or a flag. I have attached my sliceview images as well as the corresponding 3D volume images. I would really appreciate if you could take a look at them. Thank You Anisha

2 1

Question about brain masking
by Dan Dillon 06 Dec '07

06 Dec '07

Dear FreeSurfers, I'm having some trouble with my brain masks. Below I have inserted a subject's T1.mgz and brain.nii overlay (made with preproc-sess). As you can see, the mask goes well beyond the boundaries of the brain. Any ideas what could cause this-perhaps noisy functional scans? The majority of my subjects' masks look like this. Should I worry about this? My brainmask.mgz files are all fine, and I can use those to mask out voxels for visualization purposes-Do I need to be concerned if my brain.nii volumes look this way? Would I fix by running mkbrainmask-sess and tweaking the "-threshold" argument? Thanks! Dan Dillon Affective Neuroscience Lab Harvard University

2 1

estimated total intracranial volume
by Paul Greenberg 06 Dec '07

06 Dec '07

Hi All, I'm interested in the relationship between estimated total intracranial volume (eTIV), and the volume measures generated in aseg.stats and lh/rh.aparc.stats. I have used freesurfer v3.0.3 for this data set. Here I am using the output of aseg.stats for all volumes except the cerebral cortex while using the output of aparc.stats exclusively for estimates of cortical gray matter volume. As I'm looking at this data for multiple subjects I've used asegstats2table and aparcstats2table to extract the volume data for each subject (using the "--meas volume" flag). Question: When I add these volumes generated in aseg.stats (excluding cortex), and rh,lh aparc.stats (cerebral cortex only), should the total be very close to the eTIV value found in aseg.stats? Thanks, Paul -- Paul Greenberg, Ph.D. Radiology Department, University of California, San Diego VA San Diego Healthcare System (116A-13) 3350 La Jolla Village Drive San Diego, CA 92161 Phone: (858) 552-8585 ext. 2903 Fax: (858) 642-3836 pgreenberg(a)ucsd.edu

3 2

mri_convert
by Anil Roy 06 Dec '07

06 Dec '07

Hello All, I wanted to convert the brainmask volume to analyze format. I'm not very familiar with the analyze format and had two questions. 1) The wiki mentions that converting to SPM-Analyze format could involve converting to multiple or single frames. What exactly are multiple or single frames? Are they just different slices? 2) When converting, I get the message INFO: set hdr.hist.orient to 'coronal flipped'. What does that refer to? I have posted the output from mri_convert below. Thanks, Anil. ----------------------------------------------------------------------------------------------------------------------------------------------------------------- [root@slr08 freesurfer]# mri_convert -ot spm $SUBJECTS_DIR/O-06-002/mri/brainmask.mgz /root/test.img mri_convert -ot spm /mnt/win/MRI_data/MWIN4/Freesurfer//O-06-002/mri/brainmask.mgz /root/test.img $Id: mri_convert.c,v 1.146 2007/07/26 19:26:26 greve Exp $ reading from /mnt/win/MRI_data/MWIN4/Freesurfer//O-06-002/mri/brainmask.mgz... TR=2300.00, TE=3.36, TI=900.00, flip angle=9.00 i_ras = (-1, 0, 0) j_ras = (0, 0, -1) k_ras = (0, 1, 0) writing to /root/test.img... Analyze Output Matrix -1.000 0.000 0.000 129.678; 0.000 0.000 1.000 -82.220; 0.000 -1.000 0.000 140.347; 0.000 0.000 0.000 1.000; -------------------- INFO: set hdr.hist.orient to 'coronal flipped' [root@slr08 freesurfer]#

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Freesurfer