[Eeglablist] How to choose proper High-pass (low-cutoff) filter?

[Gangadhar Garipelli]

Dear Mengyan Zhu,

Selection of spectral filtering (both, selection of high-pass and 
low-pass cutoff as well as causal/non-causal ) is when analyzing EEG, is 
often overlooked. It depends on what aspect of ERP you may want to 
analyze (eg., timing of peaks, amplitudes of peaks, onset of an ERP or 
discriminant single-trial analysis).

Please refer to recent detailed debate on the topic in the following 
references:

[1] A Widmann and E Schrger. Filter effects and filter artifacts in the 
analysis of electrophysiological data. Front Psychol, 3:233, 2012.
[2] G. A. Rousselet. Does filtering preclude us from studying ERP 
time-courses? Front Psychol, 3:131,
2012.
[3] D.J. Acunzo, G. Mackenzie, and M. C. W. van Rossum. Systematic 
biases in early ERP and ERN components as a result of high-pass 
filtering. J Neurosci Methods, 209(1):212-218, Jul 2012.
[4] R. Vanrullen. Four common conceptual fallacies in mapping the time 
course of recognition. Front Psychol, 2:365, 2011.
56
[5] G. Garipelli, R. Chavarriaga, J. del R Millan, Single trial 
recognition of anticipatory slow cortical potentials: The role of 
spatio-spectral filtering, International IEEE/EMBS Conference on Neural 
Engineering 06/2011;

Wish you the best,
Gangadhar Garipelli


On 11/18/12 9:41 AM, 诸梦妍 wrote:
> Dear eeglablist,
>
> I would like to know whether there are any limitation or rules in 
> filtering EEG data, or if there are someone who could recommend some 
> papers about how high-pass filter change the data.
>
> I have read some basic principles about how filter works, but still 
> don't know how to choose a proper High-pass parameter to remove 
> artifacts without distorting the data. I knew some researchers use 
> 0.01Hz, some would recommend  0.1, 0.5 or 1Hz and some recommend not 
> using High-pass filter at all.
>
> The thing is, I found some participants' EEG waves in my experiments 
> had large amount of slow drift, which could be removed by a 0.3Hz 
> filter. However, I don't know whether that would distort the data 
> greatly. What I want to observe are P300, N400 and P600 components as 
> well as some sustained late negativity in ERP and also 3-70Hz in 
> Time-Frequency domain.
>
> Thanks for any information you may supply me with.
>
> Zhu Mengyan
> -- 
> Mengyan Zhu
> Psychology department, Peking University
> Dormitory 2061, Building 48,No.5 Yiheyuan Road, Haidian District, 
> Beijing 100871, China
>  E-mail: bj12116 at gmail.com <mailto:bj12116 at gmail.com>
>
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu

-- 
Dr. Gangadhar GARIPELLI,

Post-doc
EPFL SV BMI LNCO
SV 2805 (Batiment SV),
Station 11, 1015 Lausanne
gangadhar.garipelli at epfl.ch
http://people.epfl.ch/gangadhar.garipelli

Scientist
MindMaze, EPFL Garage
Chemin de la Raye 13
1024 Ecublen VD
Phone: +41 21 552 0805
www.mindmaze.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20121119/0a85f1a0/attachment.html>