[Eeglablist] rejecting/interpolating bad channels?

[Mahesh Casiraghi]

Thank you Arno,


I was unaware of this function, which seems indeed very useful and flexible.
But still, correct me if I am wrong, this measure is not sensitive to how
the power in a given frequency band is dinstributed across time, and across
trials.


For instance, you can have a situation1 in which 10 trials out of 100 are -
for a given channel and a given frequency band - prominently oscillating
between plus minus 700microV and the further 90 trials with an average power
within 2 SDs, and a situation2 where the same power makes the channel being
over 10 SDs in all of your trials. In the first case, once you will reject
the 10 bad epochs, you will in fact end up with a good 'preserved' original
channel, while in the second it will indeed be recommended to reject and
interpolate it.


But yes, at the end of the day, I think I will probably end up switching
from kurtosis to this more flexible and probably informative function you
propose.


Mahesh




Mahesh M. Casiraghi
PhD candidate - Cognitive Sciences
Roberto Dell'Acqua Lab, University of Padova
Pierre Jolicoeur Lab, Univesité de Montréal
mahesh.casiraghi at umontreal.ca

I have the conviction that when Physiology will be far enough advanced, the
poet, the philosopher, and the physiologist will all understand each other.
Claude Bernard




On Fri, May 13, 2011 at 12:32 AM, Arnaud Delorme <arno at ucsd.edu> wrote:

> Dear all,
>
> these days, I personally prefer to use the pop_rejchanspec function (from
> the command line only) that is based on spectrum in given frequency bands
> (several bands may be used). I would call for instance.
>
> [EEG indelec] = pop_rejchanspec( EEG, 'freqlims', [0 10; 35 128],
> 'stdthresh', [-15 15; -10 10]);
>
> This will remove data channels which have a spectral power between 0 and 10
> Hz that is more than 15 standard deviation compared to other channels and
> channels which have a spectral power between 35 and 128 Hz that is more than
> 10 standard deviation compared to other channels.
>
> Arno
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20110513/b682ba3b/attachment.html>