[Eeglablist] ASR for very low density system

[Akshay Ravindran]

Hi,

I was planning on using ASR in the Pre-processing stage for one of the
analysis I am doing and I was a bit confused with a couple of things. Would
someone help me get a better understanding of them ?


1) Can there be successful reconstruction if the system is having very low
density ? i.e. a 4 channel system by muse(2 frontal (AF7 ,AF8) and 2
temporal(TP9&TP10) channels) ? Will it be rank deficit if the channels are
just 4  especially considering the fact that many at times two channels
would be noisy?

2) Since Muse has widely separated channels, the spatial separateness
should affect the reconstruction, as it assumes correlation right? Is there
a problem in doing reconstruction in this case?

3) If clean segments are automatically found out, should I remove signals
>100uV prior to ASR if a large part of the total signal is noisy?


4) To my understanding the component space is respective of the number of
channels and no other variant exists for ASR algorithm.


Thanks
-- 

Regards,


*Akshay Sujatha Ravindran*

*Ph.D. Graduate Student | Electrical Engineering| University of Houston*

*Research Assistant | Non Invasive Brain Machine Interface Lab | University
of Houston*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170604/adb04d89/attachment.html>