[Eeglablist] Notch filter and minimum #channel for ICA

[arno]

Dear Masaki,

Masaki Nishida wrote:
> 1. I have tried to notch filter to eliminate 60Hz line noise, however, the 
> amplitude has remarkably decreased
> and the notch noise was unsuccessfully removed in the afterward specttopo().
> I am afraid that it could be a kind of bugs of band filter but we are not 
> sure.
>   
We have had problem reported with the notch filter before. The 
non-linear IIR filter works really well though and I would advise you to 
use it instead.
> 2. We have intended to apply ICA for our sleep EEG data (sampling rate: 256 
> Hz).  Although our data has enough time length for
> several hours, it has only four channels so we can get only four ICA 
> components.
> Could it be unsuitable to run ICA for our data?
>   
Yes, you might obtain interesting components although as you mentioned 
you will only obtain 4.

Best regards,

Arno