Hello all,
I am trying to keep only those EEG channels having a correlation coefficient of < 0.3 for my experiment. I have created the sample correlation between 74 channels (attached in image), but cannot select the specific subset of channels based on specific criteria.
Please help me in this regard.
Initial_corr_S1_trial2.tif (673.9 KB)
Thanks
Debashis
a correlation coefficient of < 0.3
It looks like all the channels have correlation values > 0.5 (except maybe PO7) with at least one other channels.
This is normal with EEG. An electrode that has a correlation below 0.3 with its neighbors is possibly bad (not recording any physiological signals, or with a very poor SNR).
I am trying to keep only those EEG channels
You can mark as bad the channels you want to exclude from the display and the analyses.
https://neuroimage.usc.edu/brainstorm/Tutorials/BadChannels
With a Matlab script, you can find the channels with a connectivity below or above a specific threshold (from the TF field of the connectivity file), and mark them as bad in all the data files (setting the corresponding entry in ChannelFlag to -1).
Ok..Thanks for the response. I am selecting the desired connectivity value by thresholding (Top n%) approach.