Freesurfer

freesurfer@nmr.mgh.harvard.edu

30132 discussions

options for a poor T1
by Jason Tourville 01 Feb '18

01 Feb '18

We have a few scans from kids that look like the image shown in the attached pic. We see a falloff in signal intensity in the gray matter that results in lots of cortex being excluded from the pial surface (e.g., along the dorsal surface and lateral temporal areas in the image). Typically, we would toss out this data and try to require from the subject or drop the subject from our analyses. But because of the age and clinical diagnosis of this subject we'd like to try to salvage this scan. Can you offer any suggestions for how to improve the reconstruction? Is there a parameter that we can change that will allow the pial surface to fall farther from the white matter surface in terms of voxel intensity and distance? We would trade the need for more manual edits to get greater inclusion of cortex within the pial surface. Do you have any tools or suggestions for a better intensity normalization of the original T1? Thanks for any help that you can provide. Best, Jason -- Jason A. Tourville, Ph.D. Research Assistant Professor Department of Speech, Language, and Hearing Sciences Boston University 677 Beacon St. Boston, MA 02215 Phone: (617)353-9484 Fax: (617)353-7755

2 1

Resolution difference in longitudinal analysis
by Zeynep Basgoze 01 Feb '18

01 Feb '18

Hi dear freesurfer gurus, I'm trying to do a longitudinal analysis, and whenever I run recon-all -base I face this error: "WARNING: Image geometries differ across time, maybe due to aquisition changes? This can potentially bias a longitudinal study! Will continue in 10s." I have checked the archive before writing to you. I am using freesurfer 5.3.0, and apparently this version is a little sensitive towards this issue; however even if I tried to add ' --notallow-geo ' to the code as suggested, I keep getting this warning: "Volumes differ in resolution v1res 1.000025 1.000000 1.000000 v2res 1.000099 1.000000 1.000000 diff -0.000074 0.000000 0.000000" The sequence was identical across time points, both volume & surface based outputs from recon-all look fine. I am just worried that having these resolution differences would cause us to interpret the longitudinal analysis results incorrectly. Therefore I'd like to know how seriously this much of a resolution difference would affect our results. Thank you very much in advance. Best regards, Zeynep P.S. I also attach the whole log just in case (the error I mentioned is highlighted).

1 0

Fwd: Differences between version 5.2.0 and 6.0.0
by Patrik Krumpolec 01 Feb '18

01 Feb '18

Hello! Thank You for the answer. So, If I understand this in good way - version 6.0 works with original data voxel size and the data are not conformed to voxel size 1x1x1 mm. Can I ask if also version 5.3 conform data that have different voxel size from 1x1x1 mm. Thank You very much! patrik krumpolec 2018-02-01 16:13 GMT+01:00 Bruce Fischl <fischl(a)nmr.mgh.harvard.edu>: > Hi Patrik > > we report both the number of voxels and the volume of each structure in > the stats files. Even for 1mm data we use partial volume estimation in the > newer versions so that these numbers will not be the same. In general we > find that the partial volume estimation helps accuracy, and that 6.0 is > usually more accurate than 5.2 > > cheers > Bruce > > > On Thu, 1 Feb 2018, Patrik Krumpolec wrote: > > > I want to ask about differences in brain segmentation with freesurfer > version 5.2.0 and 6.0.0.My > > dicom data have original size 0.375x0.375x0.75 mm and 600, 640, 224 > voxels. If I process the data > > vith the version 5.2.0, freesurfer conformed the data to 1 mm size and > 256 voxels for all > > directions. In the result brain structures has volume eguals to number > of voxels. > > But if i use version 6.0.0, there are data not conformed and in the > results the volume of structure > > is different form number of voxels. > > > > So, now I don't know what results is correct: from the version 5.2.0 or > 6.0.0 > > > > At first I think that version 6.0.0 is not good because my collegues > from another institution have > > results with the volume eguals to number of voxels (version 5.3) but > they have data with voxel size > > 1x1x1 mm. So, I think that it is natural that volume is eguals to number > of voxels. But my data have > > not 1x1x1 mm voxel size. > > > > Please, Could You help me! Thank You very much! > > > > patrik krumpolec > > > > > _______________________________________________ > Freesurfer mailing list > Freesurfer(a)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. > >

2 1

Re: [Freesurfer] Combining scans with different protocols
by Gudbrandsen, Maria 01 Feb '18

01 Feb '18

Dear FreeSurfer Team, Unsure whether this is the correct email for queries, apologies it not. I am wondering if it is possible to combine data set from two different protocols with different slice numbers. For example, (see attached screenshots): 1st - SAG ADNI GO ACC SPGR – 196 slices, 1.2 slice thickness, 1.2 slice gap and FoV: 270, Matrix: 256x256 2nd – ADNI MPRAGE – 166 slices - 1.2 slice thickness, 1.2 slice gap and FoV: 260, Matrix: 256x256 Many thanks, Maria

2 1

Differences between version 5.2.0 and 6.0.0
by Patrik Krumpolec 01 Feb '18

01 Feb '18

I want to ask about differences in brain segmentation with freesurfer version 5.2.0 and 6.0.0. My dicom data have original size 0.375x0.375x0.75 mm and 600, 640, 224 voxels. If I process the data vith the version 5.2.0, freesurfer conformed the data to 1 mm size and 256 voxels for all directions. In the result brain structures has volume eguals to number of voxels. But if i use version 6.0.0, there are data not conformed and in the results the volume of structure is different form number of voxels. So, now I don't know what results is correct: from the version 5.2.0 or 6.0.0 At first I think that version 6.0.0 is not good because my collegues from another institution have results with the volume eguals to number of voxels (version 5.3) but they have data with voxel size 1x1x1 mm. So, I think that it is natural that volume is eguals to number of voxels. But my data have not 1x1x1 mm voxel size. Please, Could You help me! Thank You very much! patrik krumpolec

2 1

Average template issue in M1 region
by Dr. Konstantin Weise 01 Feb '18

01 Feb '18

Dear freesurfer-team, in my current studies, I am using the make_average_subject function to generate a group template of 14 subjects. I am doing TMS motor studies, so I am primarily interested in the region around M1. Independent on the number of subjects, I experience a meshing issue right on the gyral crown of the hand knob. There are 2-3 very large triangles. I attached an example image to this mail. I also saw this issue in the average brain you shipped with the software. Do you have any idea how I can fix that? I would be very grateful, if you could help me in this concern. The same problem occured when I generated the average template for only 3 of the 14 subjects. (freesurfer version 6.0.0) Thank you very much and best regards! Konstantin ----------------------------------------------------------- Dr.-Ing Konstantin Weise Research group "Cortical Networks and Cognitive Functions" Max Planck Institute for Human Cognitive and Brain Sciences Stephanstraße 1A, 04103 Leipzig, Germany Phone: +49 341 9940-2418 Fax: +49 341 9940-2624 Email: kweise(a)cbs.mpg.de

1 0

Re: [Freesurfer] annotation to volume with custom color scheme (Yeo parcellations)
by Jingwei Li 01 Feb '18

01 Feb '18

Hi Dorit, The reason why the colors look off is that mri_aparc2aseg will automatically add 1000 for left cerebral cortex and add 2000 for right cerebral cortex. Therefore, for example, a left hemisphere voxel in network 5 will have a value of 1005 in your test.mgz. Here is my suggestion: 1. Find the indices of left/right cerebral cortex by extracting the voxels with a value of 3 or 42 in the subject's aseg. 2. Use those indices to mask out the voxels that are not cerebral cortex in your test.mgz file (i.e. set non cerebral cortical voxels to be 0). 3. For the cortical voxels, you need to subtract 1000 for those between 1000 and 2000; and subtract 2000 for those greater than 2000. Best, Jingwei

1 0

dev version
by 郑凤莲 01 Feb '18

01 Feb '18

Hi all experts, I am using freesurfer dev version for segmentation of hippocampus and amydala. I have a doubt about it. 1. In FS 6.0, I can type '203-226' in freeview for viewing subfields of hippocampus. In dev version, When I type 'freeview -v nu.mgz -v lh.hippoAmygLabels-T1.v2x.FS60.mgz:colormap=lut', there are subfields of hippocampus and amydala. How can I view them respectively, and get corresponding colors to every subfield? 2. The hippoSFvolumes.txt shows the volume of body and head of some subfields, such as CA1, CA3. But When I type 'freeview -v nu.mgz -v lh.hippoAmygLabels-T1.v2x.FS60.mgz:colormap=lut', it seems to serve the body and head as a whole. Is there any way to resolve it? Sincerely, Zheng

2 1

Fw: Postdoc fellowship at Merck
by Yendiki, Anastasia 31 Jan '18

31 Jan '18

A two-year postdoctoral fellowship in the Center for Observational and Real-World Evidence (CORE) division at Merck, applying quantitative methods to improve healthcare quality/outcomes and reduce cost: https://taleo.msd.com/careersection/merck_external_career_section/jobdetail…

1 0

Yendiki Anastasia config file
by Renew Andrade 31 Jan '18

31 Jan '18

Here is the configuration file. I think is the default by tutorial.

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