[Eeglablist] Eye Blink Sampling Rate

[Stephen Politzer-Ahles]

Hello Roma,

Why did your friend say you can't use 250 Hz? Sampling rates of 250, 500,
1000 Hz or more are standard in EEG (at least in all the research I'm aware
of). In general, sampling higher than you need is never a problem (other
than the fact that the files you get will be bigger, which is only a
problem if you have limited disk space), whereas sampling at a lower rate
than you need can be a big problem (read about the Nyquist theorem to
understand why). I would never recommend sampling at 20 Hz because this
only allows you to look at activity 10 Hz and lower; eye blinks might fall
below that threshhold (although I would not count on it---I suspect that
the frequency of peoples' blink components may vary from one person to
another, some people take bigger or longer blinks than others--and in your
own message above you suggest that eye blinks might be in 8-13 Hz, which is
already passing the 10 Hz threshhold) but a lot of cognitive activity
certainly does not.

For more information on what sampling rate means, there are introductory
texts available, e.g. Luck's "An Introduction to the Event-Related
Potential Technique".

Regarding your specific questions:

1) yes, this is the definition of sampling rate. Sampling at 20 Hz means
you acquire 20 samples per second.
2) Unless you have reason to do otherwise, use the highest reasonable
sampling rate that your hardware can handle and that is common in your area
of research (read related papers to see how they do it). You can always
downsample later. Me, for example, I always sample at 1000 Hz (with a 200
Hz online low-pass filter) because this is standard in my field, it makes
the math easy (1 sample = 1 millisecond), and the 200 Hz filter means I can
analyze data at frequencies up to around 70 Hz, which is enough to include
the cognitive signals that are interest to me (if I were interested in e.g.
high gamma activity around 90+ Hz, I would need to use a higher filter and
possibly a higher sampling rate). To reiterate: there is no drawback to
using a higher sampling rate, as long as your hardware can handle it and
you have enough disk space to save the data.
3) This means the signal associated with the blink component has a
frequency of about 8-13 Hz, i.e., if you lined up a bunch of blinks
together you could fit 8-13 of them in one second.
4) EEGLAB has a resample() function.




Stephen Politzer-Ahles
New York University, Abu Dhabi
Neuroscience of Language Lab
http://www.nyu.edu/projects/politzer-ahles/

On Mon, Aug 3, 2015 at 9:39 AM, Roma Nojan <roma.nojan at yahoo.com> wrote:

> Dear EEGLAB Members:
> I have  question related to eye blink artifact. I need to know the
> convenient number of samples that can represent the blink. I have  to know
> that since I am comparing eye blink event to normal EEG data(EEG data
> without artifacts) by using an automatic way. At first, I used the same
> sampling rate which is 250 Hz for both the blink and the normal EEG, but
> later a friend told me that you can not use the same sampling rate and said
> 20 Hz  will be enough,so my questions are:
> 1- Is using 20 Hz sampling rate for the blink means that I can not acquire
> more than 20 samples per second?
> 2-What is the sampling rate that I should use for eye blink?
> 3-What  it means for the blink to have 8-13 Hz rhythm then?
> 4-If I have to down sample the data what is the convenient way to do that ?
> I will appreciate that if you can also send me sources that can help me
> about the eye blink behavior.
> Thank you in advance
> Marwa
>
> _______________________________________________
> 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
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20150805/e214553f/attachment.html>