[Eeglablist] the correct EEG data processing pipeline

[Makoto Miyakoshi]

Dear Baris,

> One important thing is whether or not we can use the ICAs that are
> found after 1hz HPF for the raw data analysis (say data after 0.1Hz
> HPF).

Here is SCCN I've heard people say it works at least empirically.
Unfortunately I can't give you a list of papers showing validity of
this approach. So I guess this is one of do-it-at-your-own-risk
things. Personally, I see no problem in doing this.

Makoto


2012/10/28 Baris Demiral <demiral.007 at gmail.com>:
> I wonder de-trending and removal of channel mean can help.
> One important thing is whether or not we can use the ICAs that are
> found after 1hz HPF for the raw data analysis (say data after 0.1Hz
> HPF). Can we replace ICs like that? Assume that I am interested in
> very slow waves. What would be the strategy in this case?
>
> I could suggest to the initial post that cleaning the data (rejecting
> the portions with dramatic activity) should be done before ICA.
>
> Second, if you are going to use adjust, you do not need to reject the
> EOG with AAR, since adjust can correct it via ICA.
>
> Best,
> Baris
>
>
>
>
>
> On Fri, Oct 19, 2012 at 12:19 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu> wrote:
>> Dear Hui-bin,
>>
>> Yes. ICA can be easily disturbed by low frequency artifact such as
>> sweating etc. This is why we use 1Hz high-pass, which is very high.
>> You actually don't need to use this 1Hz high-passed data, because once
>> you get ICA weight you can use it for 0.1Hz high-passed data, for
>> example.
>>
>> Makoto
>>
>> 2012/10/19 Hui-bin Jia <420247417 at qq.com>:
>>> Dear Makoto,
>>>       Do you mean applying a high-pass filter before ICA will get less
>>> artifact Independent components?
>>> Thank you!
>>>
>>> Hui-bin
>>>
>>>
>>>
>>> ------------------ Original ------------------
>>> From:  "mmiyakoshi"<mmiyakoshi at ucsd.edu>;
>>> Date:  Fri, Oct 19, 2012 10:34 PM
>>> To:  "Hui-bin Jia"<420247417 at qq.com>;
>>> Cc:  "eeglablist"<eeglablist at sccn.ucsd.edu>;
>>> Subject:  Re: [Eeglablist] the correct EEG data processing pipeline
>>>
>>> Dear Hui-bin,
>>>
>>> Applying a high-pass filter helps ICA, so you may want to consider it.
>>>
>>> Makoto
>>>
>>> 2012/10/18 Hui-bin Jia <420247417 at qq.com>:
>>>> Hi, everyone
>>>>
>>>> I have been studying EEGLAB for about one year. But until now I’m not sure
>>>> about my EEG data processing pipeline is right or not.
>>>>
>>>> My EEG data processing pipeline is:
>>>>
>>>> (1) Load EEG data into EEGLAB
>>>>
>>>> (2) Read channel locations (Edit > Channel locations)
>>>>
>>>> (3) Re-reference
>>>>
>>>> (4) Use AAR ( a plugin of EEGLAB) reject EOG and EMG
>>>>
>>>> (5) Run ICA and reject artifact components using ADJUST which is also a
>>>> plugin.
>>>>
>>>> (6) Filter the data
>>>>
>>>> (7) Extract epochs and Remove baseline
>>>>
>>>> (8) Reject epochs (Tools > Reject data epochs)
>>>>
>>>> Is my method right?
>>>>
>>>> Any reply will be greatly appreciated!
>>>>
>>>> THANK YOU!
>>>>
>>>> Hui-bin Jia
>>>>
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> 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
>>>
>>>
>>>
>>> --
>>> Makoto Miyakoshi
>>> JSPS Postdoctral Fellow for Research Abroad
>>> Swartz Center for Computational Neuroscience
>>> Institute for Neural Computation, University of California San Diego
>>
>>
>>
>> --
>> Makoto Miyakoshi
>> JSPS Postdoctral Fellow for Research Abroad
>> Swartz Center for Computational Neuroscience
>> Institute for Neural Computation, University of California San Diego
>>
>> _______________________________________________
>> 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



-- 
Makoto Miyakoshi
JSPS Postdoctral Fellow for Research Abroad
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego