Freesurfer September 2012

freesurfer@nmr.mgh.harvard.edu

89 participants
154 discussions

questions about recon-all pipelines
by pietro de rossi 26 Sep '12

26 Sep '12

Dear FreeSurfers, I am a beginner, trying to process some 3T MPRAGE scans on my own for a project. I have a couple of simple questions 1. When I run the recon-all –autorecon1 –autorecon2 command it asks me for 001.mgz in mri/orig so I copy and paste the volume into the folder and the command works. My question is: do I need to change 001.mgz to 002, 003 an so on for every scan I process? 2. I have to process 3T scans in a .dcm format. Can I just process the scan as it is? Do I have to save the .dcm files as another format, maybe a .mnc file? thank you very much in advance, Pietro -- Pietro De Rossi, MD ------------------------------------- Sapienza Università di Roma, Facoltà di Medicina e Psicologia, Dipartimento NESMOS (Neuroscienze, Salute Mentale, Organi di Senso), Ospedale Sant'Andrea, Via di Grottarossa 1035-1039, 00189 Roma NESMOS Department (Neurosciences, Mental Health and Sensory Functions), School of Medicine and Psychology, Sapienza University, Sant’Andrea Hospital, Via di Grottarossa 1035-1039, 00189 Rome, Italy Neurobehavioral Clinical Research Section, Social and Behavioral Research Branch, National Human Genome Research Institute (NHGRI), National Institutes of Health (NIH), Bethesda, Maryland 20892, USA

2 1

how to make avgsubject for hippocampal subfields
by BumSeok Jeong 26 Sep '12

26 Sep '12

Dear Freesurfer expert, I'm interested in hippocampal subfields and performed it at each subject. If I'd rather get average hippocampal subfields (volume or surface), what I have to do? The make_average_subject command did not produce it. Does Freesurfer have average hippocampal subfield volume? Best wishes, Jeong --

2 1

Longitudinal processing 3 time point error
by Kawadler, Jamie 26 Sep '12

26 Sep '12

Hello, I have a dataset of patients with 3 time points each. While 2-time point processing works just fine in the longitudinal stream, when I run a base with 3 time points, the process fails and exits with errors. Please see the output below: Resolution: 0 S( 256 256 256 ) T( 256 256 256 ) Iteration(f): 1 -- diff. to prev. transform: 0.160049 Iteration(f): 2mri_robust_template(778) malloc: *** mmap(size=262144) failed (error code=12) *** error: can't allocate region *** set a breakpoint in malloc_error_break to debug MRIalloc(256, 256, 256): could not allocate 262144 bytes for 129th slice Cannot allocate memory Darwin Jamies-iMac.local 11.4.0 Darwin Kernel Version 11.4.0: Mon Apr 9 19:32:15 PDT 2012; root:xnu-1699.26.8~1/RELEASE_X86_64 x86_64 recon-all -s EB1312_61-66-77_base exited with ERRORS at Wed Sep 26 12:19:11 BST 2012 This is similar to a message on the mailing list from Liz Bowman (30 Jul 2012), in which the response was to get more memory. My command was done on a Mac with 16GB of RAM and ample disk space. I'd appreciate any help! Thanks, Jamie Kawadler

2 2

motion threshold
by Mürner-Lavanchy, Ines 26 Sep '12

26 Sep '12

Dear FreeSurfers, Due to motion I have to exclude participants from my analysis. Is there an objective measure which could serve as a cut off value concerning motion in the scanner? Can I find it in one of the FreeSurfer files generated through the recon-all process? Subjectively, one can see which images are good or good enough for segmentation. But I'd rather like to include or exclude participants based on an objective threshold. Thank you very much in advance, Best regards, Ines

1 0

Reconstructed skull malformations
by Vancelette, Reid Vincent 25 Sep '12

25 Sep '12

Hi group, I recently reconstructed a high-resolution 3D T1-weighted structural MRI using Freesurfer. The subject had non-metallic extensions that were tied to her head through a series of braids. Looking at the MRI in TkMedit, I can noticeably see pieces of the extensions above her scalp. My group performs MR-constrained MEG/EEG analyses and use the outer surface of the scalp as well as the outer and inner surfaces of the skull from the Freesurfer reconstructed MRI to calculate a forward solution. Looking at the scalp, I notice large convolutions and indentations in the reconstructed skull in an MNE program: mne_analyze. My suspicion is that the hair extensions that appear in the reconstructed MRI are causing these indentations in the reconstructed skull. Is there a way to get rid of the artifacts caused by the extensions during the reconstruction process? I am willing to provide whatever information is necessary to resolve this issue. Thanks in advance, Reid

1 0

reference of adult normal volume from cortical parcellation
by tulu 25 Sep '12

25 Sep '12

Dear FS experts, Is there any reference about the volume of cortical gyrus after FS cortical parcellation in normal adult people ? I need the reference value of regional volume base on the cortical parcellation made by FS. Please instruct me the reference paper. Thanks help JLH

1 0

Brain Volume
by Ana Arruda 25 Sep '12

25 Sep '12

Hi, I've been working with Freesurfer and FSL in order to compare the two softwares. However the last one has a tool called "Brain Extraction Tool" (BET) that provides all the brain volume including cerebellum. Is there any option in Freesurfer that makes the same as *BET*? Thank you for your attention, Ana Arruda

3 4

Pooled Structure
by Zsolt Cselényi 25 Sep '12

25 Sep '12

Hi, I would like to find out how to pool parcellation statistics (i.e. ?.aparc.stats table entries) from several areas into a bigger one. For example: how to get the pooled statistics for the area "parietal" which would be made up of "superiorparietal" and "inferiorparietal" cortical structures? Calculating pooled values for things like GrayVol, SurfArea seem rather straightforward (by summing of course) but not so for ThickAvg, ThickStd, MeanCurv, GausCurv, FoldInd and CurvInd. First, should they be pooled using the area or the number of vertices? Second, how should they be pooled? For ThickAvg and MeanCurv I guess it's an area- (or NumVert- ?) weighted mean and for ThickStd maybe a NumVert-weighted pooled SD but not sure whether it's also a (weighted) mean for GausCurv, FoldInd and CurvInd? Please help! Thank you, Zsolt

2 1

manual parcellation
by Linda Douw 25 Sep '12

25 Sep '12

Hi experts, I want to make a new, custom-made parcellation, based on the aparc but with much higher resolution. (I've tried mris_divide_parcellation, which yields strip-shaped instead of isometric labels, and mris_make_face_parcellation, which loses spatial specificity. The wiki describes a way of manually drawing new labels in tksurfer, but I get stuck in the process. I've created a new CLUT, with no. 100-106 being the new subdivisions of the original bankssts label: # Freesurfer LUT information extracted from annotation file: # lh.aparc.annot # 2008-02-29 23:58:32 # This file can be used as the basis for a new LUT file for annotations purposes # which can be loaded into tksurfer using File > Label > Load color table #No. Label Name: R G B A 0 unknown 25 5 25 0 1 bankssts 25 100 40 0 2 caudalanteriorcingulate 125 100 160 0 3 caudalmiddlefrontal 100 25 0 0 4 corpuscallosum 120 70 50 0 5 cuneus 220 20 100 0 6 entorhinal 220 20 10 0 7 fusiform 180 220 140 0 8 inferiorparietal 220 60 220 0 9 inferiortemporal 180 40 120 0 10 isthmuscingulate 140 20 140 0 11 lateraloccipital 20 30 140 0 12 lateralorbitofrontal 35 75 50 0 13 lingual 225 140 140 0 14 medialorbitofrontal 200 35 75 0 15 middletemporal 160 100 50 0 16 parahippocampal 20 220 60 0 17 paracentral 60 220 60 0 18 parsopercularis 220 180 140 0 19 parsorbitalis 20 100 50 0 20 parstriangularis 220 60 20 0 21 pericalcarine 120 100 60 0 22 postcentral 220 20 20 0 23 posteriorcingulate 220 180 220 0 24 precentral 60 20 220 0 25 precuneus 160 140 180 0 26 rostralanteriorcingulate 80 20 140 0 27 rostralmiddlefrontal 75 50 125 0 28 superiorfrontal 20 220 160 0 29 superiorparietal 20 180 140 0 30 superiortemporal 140 220 220 0 31 supramarginal 80 160 20 0 32 frontalpole 100 0 100 0 33 temporalpole 70 70 70 0 34 transversetemporal 150 150 200 0 100 bankssts_1 25 100 41 0 101 bankssts_2 25 100 45 0 102 bankssts_3 25 100 50 0 103 bankssts_4 25 100 55 0 104 bankssts_5 25 100 60 0 105 bankssts_6 25 100 65 0 I load the aparc and the new CLUT into tksurfer, draw points around a new label, close the path, and then custom fill the area. How do I now indicate that this is for instance the new label bankssts_1? The wiki and mailing archive indicate some problems with manually parcellating, but the info seems dated and scattered. So if there is some place I can find this stuff myself, I'd be happy to check it out. Thanks so much, -- Linda Douw, Ph.D. Research fellow

1 0

recon-all error while data's FOV>256
by bin.lv 25 Sep '12

25 Sep '12

Dear all, I met a problem while running the recon-all. I have a data with FOV= 300>256, so I add the flag -cw256 with reon-all. Then when I check the converted data raw.mgz, I find the brain image is imperfect while some region at the top of head is cut off. My question is, could I choose to cut the part of neck which locates at the button of head? Thank you very much in advance！ Bin Lv

1 0

← Newer
1
2
3
4
5
6
7
8
...
16
Older →

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 September 2012