[Eeglablist] With Respect to Removing AAR
khosr023 at umn.edu
Mon Jun 26 17:36:04 PDT 2017
I had a question with respect to the removal of EMG artifacts from my EEG
data. I had planned to apply AAR after the application of Cleanline on my
continuous EEG (before re-referencing to the average of all channels and
before epoching my channels). But sometimes I receive the error that "AAR
needs exactly three outputs" and then the algorithm terminates. I
downloaded it again, and did not realize how, but at least so far, it is no
more there. But I would appreciate any comments on how to deal with this
issue if it occurs again.
My second question was about the dimensionality of my ICA matrix. When I
apply the pre-ICA pre-processing (cleanline, AAR, removing bad channels and
re-referencing to average), at the beginning of running the ICA I get this
message that the rank of the matrix is reduced. This deprives me from being
able to apply the ADJUST algorithm on my ICs for artifact removal, since it
requires full rank matrices (64 components for a 64-channel recording).
When it happened, I decided to use MARA instead. I would very appreciate
any comments and ideas in this respect, and if there are any pros and cons
with either of these approaches. One of the main sources of noise that I
would need to remove is the EMG noise.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the eeglablist