[Mne_analysis] mne_volume_source_space with --mri option: segmentation fault

[Matti Hamalainen]

Hi,

The unfortunate answer is that the interpolator cannot be created for discrete source spaces. I changed the code to prevent such an attempt and the fixed software will be in the nightly build tomorrow.

Best,
Matti
 
On Jan 7, 2014, at 5:21 PM, pavitrak at nmr.mgh.harvard.edu wrote:

> Hi all,
> 
> I am trying to create a volume source space using a text file
> containing the source locations.  However, I get a segmentation fault
> (core dumped) error if I use the mri option.  The position file is in
> MRI (surface RAS) coordinates as recommended by the MNE manual.  Here
> is an example of the error -- this kind of error occurs regardless of
> what text file or MRI volume (T1/brain/brainmask etc) I give it.
> 
> ----------------------------------------------------------------------------
> mne_volume_source_space --pos
> $SUBJECTS_DIR/$SUBJECT/src/subject_lmth.txt --mri T1.mgz --src
> $SUBJECTS_DIR/$SUBJECT/vol-src.fif
> 
> mne_volume_source_space version 1.9 compiled at Dec 15 2009 08:55:40
> 
> Source location file  :
> /autofs/cluster/purdonlab/users/pavitrak/sourceloc/Validation_Data/AEP/cparty_001/src/subject_lmth.txt
> Assuming input in millimeters
> Assuming input in MRI coordinates
> Output file           :
> /autofs/cluster/purdonlab/users/pavitrak/sourceloc/Validation_Data/AEP/cparty_001/bem/vol-src.fif
> 
> Positions only (in millimeters)
> 150 sources
> Reading T1.mgz...[done]
> Setting up interpolation...Segmentation fault (core dumped)
> ----------------------------------------------------------------------------
> 
> If I exclude the mri option the source-space file is created just fine
> -- so perhaps its a fault to do with the interpolation into the voxel
> space of the MRI volume.  But I need the interpolator matrix as I plan
> to use mne_volume_data2mri to create an overlay of time series
> estimates corresponding to these volume source spaces.  Why is this
> happening and is there any way to overcome/circumvent this problem?
> 
> Thanks so much.
> Best Regards,
> Pavitra
> _______________________________________________
> Mne_analysis mailing list
> Mne_analysis at nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/mne_analysis
> 
>