Can we filter alpha wave for the ERP when alpha wave is dominant?

Hello Mike,

I want to see the evoked potential from my data. The subject was asked to close his eyes and the stimuli was given at eyes-closed condition. When I calculated the evoked potential, Alpha wave is covering the evoked potential. How can I overcome this issue?

Thank you so much

Hi Earl. Interesting question. It depends on whether you consider alpha to be a real part of the stimulis response, or an “artifact” of the eyes-closed response. If the former, then the ERP is still valid, or you can use a time-frequency analysis to separate the alpha activity from activity at other frequencies (e.g., theta or gamma).

If you consider the alpha to be a nuisance and you want to get rid of it, then you’ll need to filter it out. You can do that by low-pass filtering at, e.g., 8 Hz, or by applying a narrowband stop filter from, e.g., 8-12 Hz. To be honest, I find this solution to be a bit strange, because alpha is also involved in stimulus processing. But from a signal-processing perspective, this is certainly a valid strategy.

1 Like

Thank you so much Mike, filtering at 8 Hz works fine.
Will researchers accept this results by not filtering alpha. I am applying this for N2P2 potential. Most references use bypass filter 1-200 Hz.

So do you accept this kind of strategy when not applying alpha.

Thank you so much.

I don’t do ERP research, so I’m not well-versed in what’s deemed normal or acceptable. But in my opinion, the important thing is to explain your analyses and justify your choices. You could also show the data with and without the filtering, so people can see the effect.

1 Like