[Mne_analysis] mne_do_inverse_operator

[William W. Graves]

Dear MNE Experts,

I suspect this is a question for Matti, but of course I'll be happy to get
help from anyone. When I launch mne_do_inverse_operator, it goes through
the whole process, then at the very end it gives a "Segmentation fault".
Oddly, I have a group of 20 participants, all analyzed the same way, but
it only does this for one of them. I'd appreciate any help/insight you can
give. Here's the full screen output, in case it helps:

wgraves at argus:~/data/ld_fmri/data/4803/freesurf/4803/MEG/run03_lexdec01>
mne_do_inverse_operator --fwd $megdir/$rundir/${meas}-ico-4-fwd.fif
--fixed --depth --megreg 0.1 --noisecov $megdir/$rundir/$cov

mne_inverse_operator version 2.33 compiled at Aug 14 2013 04:19:05

Compute the MNE inverse operator decomposition

Forward solution                          :
/home/wgraves/data/ld_fmri/data/4803/freesurf/4803/MEG/run03_lexdec01/run03_xtraClean_0.1_40Hz_word_nonword_ave-ico-4-fwd.fif
Noise covariance matrix                   :
/home/wgraves/data/ld_fmri/data/4803/freesurf/4803/MEG/run03_lexdec01/run03_xtraClean_0.1_40Hz_word_nonword-cov.fif
Source covariance matrix                  : identity matrix
Destination for the inverse operator data :
/home/wgraves/data/ld_fmri/data/4803/freesurf/4803/MEG/run03_lexdec01/run03_xtraClean_0.1_40Hz_word_nonword_ave-ico-4-meg-fixed-inv.fif
Include MEG data.
Noise-covariance regularization (mag)     : 0.10
Noise-covariance regularization (grad)    : 0.10
Employ depth weighting.
        limit    = 10.00
        exponent = 0.80

Reading the forward solution....
        Read data for 306 MEG channels and 5112 sources
        Free source orientations.
        The forward computation was performed in head coordinates.
        Read 2 source spaces with a total of 5112 source locations
        Source spaces are now in head coordinates.
        Channel description list matched with the composite forward
solution matrix.
        Local surface coordinate system forward matrix will be employed.

Loaded projection from
/home/wgraves/data/ld_fmri/data/4803/freesurf/4803/MEG/run03_lexdec01/run03_xtraClean_0.1_40Hz_word_nonword-cov.fif:
        # 1 : planar-999--0.080-0.080-PCA-01 : 1 vecs : 204 chs MEG active
        # 2 : axial-999--0.080-0.080-PCA-01 : 1 vecs : 102 chs MEG active
        # 3 : planar-801--0.200-0.200-PCA-01 : 1 vecs : 204 chs MEG active
        # 4 : axial-801--0.200-0.200-PCA-01 : 1 vecs : 102 chs MEG active

Read a full noise covariance matrix from
/home/wgraves/data/ld_fmri/data/4803/freesurf/4803/MEG/run03_lexdec01/run03_xtraClean_0.1_40Hz_word_nonword-cov.fif
Picked appropriate channels from the sensor noise covariance matrix.
MEG/EEG correlations omitted.
Average noise-covariance matrix diagonals:
        Magnetometers       : 170.14  fT    reg = 0.10
        Planar gradiometers : 38.99   fT/cm reg = 0.10
Noise-covariance regularized as requested.
Projection applied to the covariance matrix.
Decomposing the noise covariance...
        Estimated covariance matrix rank = 302 (0.00206112)
        302 MEG and 0 EEG-like channels remain in the whitened data
done.

Creating the depth weighting matrix...
        204 planar channels
        limit = 3613/5112 = 10.000623
        scale = 2.06846e-07 exp = 0.8
        Picked elements from a free-orientation depth-weighting prior into
the fixed-orientation one.
        Converted the forward solution into the fixed-orientation mode.
Creating the source covariance matrix:
        Depth-weighting applied.
done

Applying linear projection to the forward solution...done
Whitening the forward solution...done
Scaling the source covariance...done
Decomposing...
        Applying a priori source weighting to the forward solution...done
        Transpose...done
        SVD...done
        largest singular value = 0.171679
        scaling factor to adjust the trace = 25.2724 (nchan = 306 nzero = 4)
done

Writing the solution to
/home/wgraves/data/ld_fmri/data/4803/freesurf/4803/MEG/run03_lexdec01/run03_xtraClean_0.1_40Hz_word_nonword_ave-ico-4-meg-fixed-inv.fif...Segmentation
fault
Inverse operator decomposition failed (see above)

Thanks in advance,
Will

---
William W. Graves, Ph.D.
Assistant Professor, Department of Psychology
Rutgers, The State University of New Jersey
Smith Hall, Room 337
101 Warren Street, Newark, NJ  07102
Voice: 973-353-3947
Fax: 973-353-1171
http://lbbil.rutgers.edu