New subject: Longitudinal LME matlab error - Matrix singular or badly scaled

24 Nov 2023


      External Email - Use Caution
&nbsp;
Hello Sophie,
Just FYI, there is no matlab distribution included in the 7.4.1 release 
and we so not support using R2023b.&nbsp; So I presume someone at your site 
installed R2023b separately via a Mathworks matlab installer.&nbsp; We do provide a script, fs_install_mcr, to install either the R2014b or R2019b matlab runtime environment 
into the freesurfer tree for use.&nbsp; I don't know if running R2019b would affect what you report below, but you could try it via the commands below..
- R.
$ export FREESURFER_HOME=/usr/local/freesurfer/7.4.1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &lt;--- linux install path
$ source $FREESURFER_HOME/SetUpFreeSurfer.sh
&lt;... output showing SUBJECTS_DIR and other environment variables&gt;
$ cd &nbsp;$FREESURFER_HOME/bin
$ sudo FREESURFER_HOME=$FREESURFER_HOME &nbsp;./fs_install_mcr &nbsp;R2019b
I
 believe this sets things up so the libmwm* libraries are installed 
under $FREESURFER_HOME/MCRv97/bin/glnxa64.&nbsp; You can try adding that path
 (or any path the script output recommends) to the start of 
LD_LIBRARY_PATH.&nbsp; You should remove any other entries to a different 
MATLAB distribution.
On 11/23/23 20:42, Sophie Matis (sophie.matis@sydney.edu.au) wrote:
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;External Email - Use Caution&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Hello FreeSurfer experts!
&nbsp;
I have carried out some analysis of changes in 
cortical thickness over time using the LME Matlab (R23b) Toolbox 
distributed within Freesurfer (7.4.1) in line with the instructions on 
this page: &nbsp;MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be https://secure-web.cisco.com/16TbMFKCSxY9QLMDkUoKiHpONPO8cJ5IumheTtc9oNl4e80...
&nbsp;
At this stage I just want to look at changes over 
time within each of three groups. For two groups I have been able to 
proceed through the steps fine and have my significance files which look
 normal. For the last group I am running into
 an error on the “[lhTh0,lhRe] = lme_mass_fit_EMinit(X,[1 
2],Y,ni,lhcortex,3); “ step. MATLAB returns “Warning: Matrix is 
singular, close to singular or badly scaled” and lhTh0 and lhRe are 
created as matrices with only NaN values.
&nbsp;
Has anyone run into this error before or have any tips for troubleshooting it?
&nbsp;
I haven’t had any errors on the longitudinal 
processing steps and visual QC on my files looks normal. No errors on 
mris_preproc or mri_surf2surf. &nbsp;An aparcstats2table for thickness 
doesn’t reveal any NaNs.
&nbsp;
My design matrix is created with X = [ones(length(M),1) M] after following the suggested sortData code from the wiki
Contrast is [0 1]
This is my qdec file:
&nbsp;
fsid&nbsp;&nbsp;&nbsp; fsid-base&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; yearsfromBL
108_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 108&nbsp;&nbsp; 0.00
199_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 199&nbsp;&nbsp; 0.00
199_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 199&nbsp;&nbsp; 1.10
199_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 199&nbsp;&nbsp; 2.25
202_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 202&nbsp;&nbsp; 0.00
204_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 204&nbsp;&nbsp; 0.00
204_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 204&nbsp;&nbsp; 1.16
204_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 204&nbsp;&nbsp; 2.30
223_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 223&nbsp;&nbsp; 0.00
0821_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0821&nbsp; 0.00
0931_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 0931&nbsp; 0.00
1010_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1010&nbsp; 0.00
1010_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1010&nbsp; 1.84
1031_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1031&nbsp; 0.00
1031_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1031&nbsp; 1.00
1031_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1031&nbsp; 2.13
1064_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1064&nbsp; 0.00
1064_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1064&nbsp; 2.01
1125_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1125&nbsp; 0.00
1134_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1134&nbsp; 0.00
1134_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1134&nbsp; 1.17
1134_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1134&nbsp; 2.28
1233_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1233&nbsp; 0.00
1233_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1233&nbsp; 1.05
1373_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1373&nbsp; 0.00
1439_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1439&nbsp; 0.00
1439_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1439&nbsp; 1.14
1439_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1439&nbsp; 2.28
1475_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1475&nbsp; 0.00
1475_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1475&nbsp; 1.04
1475_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1475&nbsp; 2.03
1522_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1522&nbsp; 0.00
1522_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1522&nbsp; 2.07
1541_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1541&nbsp; 0.00
1541_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1541&nbsp; 1.47
1551_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1551&nbsp; 0.00
1551_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1551&nbsp; 0.98
1551_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1551&nbsp; 2.05
1572_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1572&nbsp; 0.00
1572_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1572&nbsp; 1.03
1625_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1625&nbsp; 0.00
1625_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1625&nbsp; 1.06
1637_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1637&nbsp; 0.00
1641_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1641&nbsp; 0.00
1656_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1656&nbsp; 0.00
1656_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1656&nbsp; 0.94
1656_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1656&nbsp; 2.26
1755_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1755&nbsp; 0.00
1755_MR2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1755&nbsp; 1.04
1755_MR3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1755&nbsp; 2.01
1829_MR1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1829&nbsp; 0.00
09154_MR1&nbsp;&nbsp;&nbsp;&nbsp; 09154 0.00
09154_MR2&nbsp;&nbsp;&nbsp;&nbsp; 09154 1.01
09154_MR3&nbsp;&nbsp;&nbsp;&nbsp; 09154 2.01
12110_MR1&nbsp;&nbsp;&nbsp;&nbsp; 12110 0.00
12110_MR2&nbsp;&nbsp;&nbsp;&nbsp; 12110 1.04
&nbsp;
&nbsp;
Any help appreciated!
&nbsp;
Thank you,
&nbsp;
Sophie Matis&nbsp;|
 Data Management &amp; Research Officer
School of Health Sciences &amp; Brain and Mind Centre&nbsp;
Faculty of Medicine and Health&nbsp;
&nbsp;
The University of Sydney&nbsp;&nbsp;
Brain and Mind Centre, 94-100 Mallett Street, Camperdown (M02G)&nbsp;|
 The University of Sydney | NSW | 06&nbsp;
sophie.matis@sydney.edu.au&nbsp;
 |&nbsp;sydney.edu.au
&nbsp;
CRICOS 00026A
This email plus any attachments to it are confidential. Any unauthorised
 use is strictly prohibited. If you receive this email in error, please delete it and any attachments.
&nbsp;
Please think of our environment and only print this email if necessary.
&nbsp;
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://secure-web.cisco.com/1JpB2M6srxvOJofQQKW2x6xhujK5Mdwvp75j2I-8SrFl7MO...
Please note that this e-mail is not secure (encrypted).  If you do not wish to continue communication over unencrypted e-mail, please notify the sender of this message immediately.  Continuing to send or respond to e-mail after receiving this message means you understand and accept this risk and wish to continue to communicate over unencrypted e-mail.

Re: [Freesurfer] Longitudinal LME matlab error - Matrix singular or badly scaled