problem with automatic artifact detection

[Elisabeth May]

Dear Fieldtrip users,

I am trying to use Fieldtrip's automatic artifact detection on a new MEG
dataset (recorded with the Neuromag 306 system) and encountered a
problem that I didn't have before. During EOG artifact detection, I get
the following warning for each of my (300) trials:

".Reading 599676 ... 606076  =    299.838 ...   303.038 secs... [done]
Warning: Matrix is close to singular or badly scaled.
         Results may be inaccurate. RCOND = 7.730460e-17.
 > In filtfilt at 81
  In filtfilt at 49
  In preproc_bandpassfilter at 73
  In fieldtrip-20100208/private/preproc at 256
  In ft_artifact_zvalue at 149
  In artifact_zvalue at 17
  In ft_artifact_eog at 121
Warning: Matrix is close to singular or badly scaled.
         Results may be inaccurate. RCOND = 7.730460e-17.
 > In filtfilt at 81
  In filtfilt at 49
  In preproc_bandpassfilter at 73
  In fieldtrip-20100208/private/preproc at 256
  In ft_artifact_zvalue at 149
  In artifact_zvalue at 17
  In ft_artifact_eog at 121"

The function nevertheless runs through but the resulting z-scores don't
make sense (see the figure bad z-scores attached to this email for an
example of a trial). I have another dataset from the same recording
session with the same subject where a different paradigm was used. With
that paradigm, the artifact detection works fine (see figure good
z-scores). This was the same for another subject who did both of the
paradigms within one recording session. The only thing I could think of
that could be important and is different between the two paradigms /
datasets is the sampling frequency; the artifact detection seems to work
fine for a sampling frequency of 1000 Hz but not for a sampling
frequency of 2000 Hz.

Since the warning refers to preproc_bandpassfilter,  I tried to track
the steps of the filtering and plotted the data of a single trial and
one EOG channel before and after the application of the bandpass filter
during the EOG artifact detection routine. I did this for both the
paradigm that results in "normal" z-scores (figures good before
filtering and good after filtering) and for the one that results in the
z-scores that don't make sense (figures bad before filtering and bad
after filtering). I don't know very much about filters, but is it
possible that the settings for the filters are somehow not working /
calculated wrongly for the 2000 Hz sampling frequency? Or am I
completely on the wrong track? Has anyone else encountered this problem
before?

Thanks in advance for any help!

Best,
Elisabeth


--
Dipl.-Psych. Elisabeth May

Universitätsklinikum Düsseldorf
Institut für Klinische Neurowissenschaften und Medizinische Psychologie
Universitätsstr. 1
40225 Düsseldorf

Tel: +49 211 81-18075

http://www.uniklinik-duesseldorf.de/med-psychologie


----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/neuroimaging/fieldtrip.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bad after filtering.jpg
Type: image/jpeg
Size: 13816 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20100416/2cd235cc/attachment-0006.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bad before filtering.jpg
Type: image/jpeg
Size: 29714 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20100416/2cd235cc/attachment-0007.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bad z-scores.jpg
Type: image/jpeg
Size: 67206 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20100416/2cd235cc/attachment-0008.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: good after filtering.jpg
Type: image/jpeg
Size: 18819 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20100416/2cd235cc/attachment-0009.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: good before filtering.jpg
Type: image/jpeg
Size: 22162 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20100416/2cd235cc/attachment-0010.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: good z-scores.jpg
Type: image/jpeg
Size: 63844 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20100416/2cd235cc/attachment-0011.jpg>