Baseline Correctiion

EEG
#1

Hi, I have a few queries regarding baseline:

  1. What is the impact of bad channel on linear baseline correction?

  2. Is it necessary to do avg re-referencing the data before applying ICA on it?

  3. Is it necessary to have baseline normalization before applying ICA so that the impact of 1/f phenomena can be reduced?

#2
  1. Subtracting noise from noise is still noise. Better to remove that electrode.
  2. No, but often recommended.
  3. ICA should be run on the time-domain data for cleaning; 1/f is a frequency-domain phenomenon that is relevant for the final step of analyses.