[Eeglablist] Running ICA with 128 channels

[Downs, Heather B]

Hi all,

We are attempting to remove the blink artifacts from our EEG recordings. Our process has included visually cleaning the re-referenced and filtered data then running ICA and removing the blink components. However, on some of the recordings we are finding that the ICA after the visual screening is less 'stable' than when we ran ICA before the visual screening. It seems that sometimes the blink component will be split between four or so components. We are using 128 channels and our recordings are about 5 minutes long before the visual screening. The shortest recording we end up with after screening is right over 2 minutes. I have been reading that this is not enough recording time to properly run ICA with 128 channels. Would this explain the issue we are encountering when we run ICA after visually screening the data? If this is the case, what would your recommendations be? We unfortunately cannot make our recordings longer as we are no longer gathering data, so I know that is not an option we have.

Thank you!

Heather Downs

----------------------------------------------------------------------
Confidentiality Notice: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170530/899f029b/attachment.html>