CMC frequencies

Hello,

I am looking to calculate coherence between EEG and EMG data in brainstorm. I am planning to do this through 1xN coherence. I am interested in the frequencies between 13 and 50Hz.

I am wondering is there a way to define a minimum frequency for the computation? So far I have been only able to compute coherence for the full range between 0 and 50Hz. Is the computation affected by also including those lower frequencies that I am not interested in or can I just ignore those.

Also, I am looking to attach the raw data to the EEG data as an extra channel to do this calculation but I have also seen people doing CMC first filter the data (e.g. bandpass filter between 10 and 400Hz). If I am only interested in the range between 13 and 50 Hz, would such a filter still be necessary? Would low or high frequency noise outside of this range impact the coherence calculation for these mid frequencies?

Thanks!
Christian S

The coherence is computed for the full spectrum anyway. The values at lower frequencies have no impact on the values at higher frequencies, you can safely ignore them. If you really want to remove them from the graphs displayed in Brainstorm, you could edit the coherence files and remove manually the unwanted frequencies. But this is not a trivial modification, and would not have a different result than cutting the left part of the spectrum with an image editor.

I am looking to attach the raw data to the EEG data as an extra channel to do this calculation

This is not easy to do this in Brainstorm. Unless you absolutely need this in order to export the combined signals to another program, I'd recommend you process the two files separately.

If I am only interested in the range between 13 and 50 Hz, would such a filter still be necessary?

Probably not. The least amount of preprocessing is always to be preferred.

Would low or high frequency noise outside of this range impact the coherence calculation for these mid frequencies?

Not much. The filter itself, on the other hand, may have a significant impact on the result.

Have you followed this tutorial?
https://neuroimage.usc.edu/brainstorm/Tutorials/CorticomuscularCoherence