[Eeglablist] ERPLAB filter

[Steven J Luck]

Dear HanGue,

I would suggest being very cautious about this approach.  Whenever a narrow passband filter is applied to a waveform, non-oscillating activity will be converted into oscillations.  The reason is that transient changes in voltage (e.g., a P3 waveform, sudden voltage shifts arising from movement artifacts, etc.) contain power across a wide range of frequencies, and filtering the data pulls out the low frequencies and turns them into oscillations.  For more information, see the chapter in filtering in my book on ERP methods (An Introduction to the Event-Related Potential Technique, MIT Press).  For a good example of the application of these ideas, see Yeung, N., Bogacz, R., Holroyd, C. B., Nieuwenhuis, S., & Cohen, J. D. (2007). Theta phase resetting and the error-related negativity. Psychophysiology, 44, 39-49.

In addition, skin potentials contain a lot of power at very low frequencies, and the oscillations you are seeing could be coming from the skin rather than from the brain (see the chapter on artifacts in my book).

Steve Luck


From: HanGue Jo <johangue at gmail.com<mailto:johangue at gmail.com>>
Subject: [Eeglablist] ERPLAB filter
Date: December 20, 2012 6:30:07 AM PST
To: <eeglablist at sccn.ucsd.edu<mailto:eeglablist at sccn.ucsd.edu>>


Dear all,

I'm interesting in the very slow oscillation of EEG, from 0.01 to 0.2Hz.
For that reason I used IIR Butterworth filter from ERPLAB twice, first high-pass (0.01, 24dB/oct) and second low-pass (0.2, 24dB/oct). Not at the same time.
the time length is more than 10 minutes in 1000Hz sampling.
In visual inspection, filttered data shows sybchronized phase with raw data. but I'm not sure about this method.
Could you give me any comment?

Best
HanGue.










-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20121221/d00498f6/attachment.html>