Averaging with mne_process_raw: zero-ed trials included?

✉️ Mailing List Archive (read-only)
1

Hello,

    I am working with MEG data collected on a Elekta Neuromag scanner and
processed with the maxfilter utility. Segments of my data become zero-ed out
in the maxtiler output, which I believe may be due to head motion beyond
some threshold where the motion correction feature no longer functions.
Post-maxfilter I begin processing with the MNE utilities. I would like to
avoid any zero-valued trials being included in an EMF or covariance matrix
created using mne_process_raw. Does mne_process_raw include any detection
for such trials, or does it include all trials which it is given? (For
instance, if all MEG channels are fixed at zero for the entire duration of
the trial.)

    Thanks again,

Per Lysne
University of New Mexico/Mind Research Network
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20110818/e5b76bb6/attachment.html

2

You should be able to use the options --magFlat and --gradFlat. Check
the MNE manual for these entries in Chapter 4---version 2.5 didn't
have it, but 2.7 (or above) does.

Eric

3

Thanks for the tip, Eric. It looks like these might be fields in the average
and covariance definition files instead of command line options? In either
case, they are exactly what I was looking for.

-Per