[Eeglablist] Preprocessing Pipeline for Artifact Subspace Reconstruction

[Makoto Miyakoshi]

Dear Srinivas,

> Highpass filter the data at 1HZ . Do I need to do this if I use
clean_rawdata?

clean_rawdata() has a high-pass filter so if you like how it works you
don't need to run eegfiltnew().

> should I use clean_artifacts with selected options? Any recommendations?

Try it with several parameter combinations until you feel confident with
it. As I wrote, in our unpublished study ASR with SD==20 showed the best.
Compare SD==20 and SD==4 to see how the results differ.

Makoto

On Thu, Nov 13, 2014 at 3:07 PM, Srinivas Kota <svkota at gmail.com> wrote:

> Hello Christian and Makoto:
>
> Would pipeline for preprocessing would be
>
> Import Data and channel information
> Highpass filter the data at 1HZ . Do I need to do this if I use
> clean_rawdata? or should I use clean_artifacts with selected options? Any
> recommendations?
>
> Srinivas
>
> On Thu, Nov 13, 2014 at 9:07 AM, Srinivas Kota <svkota at gmail.com> wrote:
>
>> Hello Tarik:
>>
>> Thanks for the detailed information. The pipeline for ASR would be
>>
>> import data
>>  remove line noise using clean line
>> High-pass filter the data at 1Hz (Do we need to do this, because ASR
>> includes a high pass filter arg_highpass?)
>> Import channel information
>> ASR
>> Remove bad channels (channel replacement if necessary?)
>> Re-reference the data to average
>> Epoch data
>> Reject epochs for cleaning
>>
>>
>> Best
>> Srinivas
>>
>>
>>
>>
>>
>> On Wed, Nov 12, 2014 at 8:52 PM, Tarik S Bel-Bahar <
>> tarikbelbahar at gmail.com> wrote:
>>
>>> Hello Srinivas, hoping the below is of use to you, cheers!
>>>
>>> ASR is a cool tool! You may use it by installing it in eeglab as a
>>> plugin, running it from the gui on a continuous EEG file, and then typing
>>> eegh to get the eeglab history. Makoto might have something to add as he
>>> made the page, but the below should be useful to you.
>>>
>>> You can also find out more at the powerpoint from Christian and Makoto
>>> at:
>>> http://sccn.ucsd.edu/wiki/EEGLAB_Extensions
>>>
>>> You can also download the matlab functions from there, and review the
>>> notes and examples given within the .m files. There is considerable
>>> information there.
>>>
>>> The change in the preprocessing pipeline suggested on the webpage would
>>> likely mean implementing ASR after importing channels.Because it removes
>>> bad channels, removes bad data, and gives you cleaned data, it's likely you
>>> can skip the other bad-channel and artifact-rejection steps. However, be
>>> sure to review the data for yourself after ASR, as it may require more
>>> cleaning, or adjustment of the parameters and running a second or different
>>> pass of artifact rejection using the same or other techniques. (however,
>>> the defaults work quite well for the methods mentioned at the very end of
>>> the pipeline webpage).
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Tue, Nov 11, 2014 at 6:48 PM, Srinivas Kota <svkota at gmail.com> wrote:
>>>
>>>> I found this processing pipeline in EEGLAB wiki page.
>>>>
>>>> http://sccn.ucsd.edu/wiki/Preprocess_pipeline
>>>>
>>>> What would be the preprocessing pipeline, if I like to try ASR?
>>>>
>>>>
>>>>
>>>> Best,
>>>> Srinivas
>>>>
>>>> _______________________________________________
>>>> 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
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20141116/d931e2a7/attachment.html>