[Eeglablist] Methods of improving SNR

[Makoto Miyakoshi]

Dear Arun,

The point in ASR 20 is to let ICA learn artifacts well but at the same time
protect ICA from outliers in the raw data.

> I have been using default values with good results. But some of our lab's
meditation data were losing visible 'alpha' patterns with this.

Right, it happened to me too. That's why I increased the threshold far
higher.

After all, you cannot completely separate and remove artifacts only. The
merit of ASR is evaluated in a total SNR increase, and there is no
guarantee that the signal will be absolutely protected e.g. if Method A
removes 10 artifacts and 1 signal, and Method B removes 30 artifacts and 2
signal, Method B is better in total.

Makoto

On Mon, Sep 29, 2014 at 9:37 AM, Arun Sasidharan <arunsasi84 at gmail.com>
wrote:

> Dear Makoto,
> Thanks for the reply. I was wondering how such a large std value would be
> used.
> I have been using default values with good results. But some of our lab's
> meditation data were losing visible 'alpha' patterns with this.
> I will use this and give my feed back.
> Once again thanks....
>
> On 29 September 2014 21:39, Makoto Miyakoshi <mmiyakoshi at ucsd.edu> wrote:
>
>> Dear Arun,
>>
>> Read the GUI carefully and you'll find one of the boxes says '... using
>> ASR' in which the default value could be 4 or 5 but change it to 20.
>>
>> Makoto
>>
>> On Sun, Sep 28, 2014 at 2:24 AM, Arun Sasidharan <arunsasi84 at gmail.com>
>> wrote:
>>
>>> Dear Makoto,
>>> In the last email with this topic ('Methods of improving SNR'), you
>>> mentioned that ASR threshold of "20" produced good results. Could you
>>> kindly help me identify the parameter of clean_raw/clean_asr that I need to
>>> change to acheve this.
>>>
>>> Thanks...
>>>
>>> --
>>> Dr Arun Sasidharan
>>> PhD Scholar & Senior Research Fellow (ICMR)
>>> Dept of Neurophysiology & Multi-modal Brain Image Analysis Laboratory
>>> (MBIAL)
>>> NIMHANS, Bangalore-560029, India.
>>> http://nphy.weebly.com/arun-sasidharan.html
>>> http://mbial.weebly.com/arun-sasidharan.html
>>> http://in.linkedin.com/pub/dr-arun-sasidharan/18/a15/352
>>> Mob: +91-9964923983
>>>
>>> _______________________________________________
>>> 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
>>
>
>
>
> --
> Dr Arun Sasidharan
> PhD Scholar & Senior Research Fellow (ICMR)
> Dept of Neurophysiology & Multi-modal Brain Image Analysis Laboratory
> (MBIAL)
> NIMHANS, Bangalore-560029, India.
> http://nphy.weebly.com/arun-sasidharan.html
> http://mbial.weebly.com/arun-sasidharan.html
> http://in.linkedin.com/pub/dr-arun-sasidharan/18/a15/352
> Mob: +91-9964923983
>



-- 
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/20140929/30adb6ee/attachment.html>