[Eeglablist] very high residual variance on clean continuous data

[Kelsey Sundby]

Hello,

I am having an issue with getting very high residual variance values
(almost no IC's <15% in all subjects). The continuous data looks pretty
clean but I manually remove bad stretches, I remove bad channels, and use
recommended filtering prior to ICA (i.e. high pass filter etc.). Is there
any recommended way to resolve this issue? We have around ~36 minutes of 64
channel data recorded for each subject which I know is on the lower side
but we thought with at least 30minutes it would be sufficient. I have been
using IC label to cross validate what I think is "brain" and IC label has
been identifying brain components but again those have very high RV. Any
recommendations/help would be very appreciated!

Thank you
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20181120/5bfbfb34/attachment.html>