Hey everyone,
I have data where the first few participants are missing some of the sensors that were bad in their preprocessed data. So, that might be the source of this issue as well. I have taken the preprocessed data and inserted it into an earlier iteration of the data that has those sensors, then I interpolate them. The issue is those sensors seem to ruin some of the smoothness of the topography. I don’t know if this is normal for interpolation, it might be. The overall maps seem stable, just those few sensors seem to add in these odd polarity holes. Attached, the first picture is the first epoch, a head model for each few milliseconds with the bad sensors missing. Note the smoothness compared to the second pic where there are some holes where there are bad sensors I interpolated. Particularly noticeable is sec1.34 where the left side sensor interpolated back in takes a bite out of the left negative polarity, and sec1.38 where dead center of the head a positive hole pops in.
For reference, this is what happens when I port over the preprocessed data but do not interpolate. Some of the channels from the ones above are declared bad here too, like the two sensors referenced above.
However, once I interpolate, I create that second pic. I tried to declare those same channels as bad again to see if it would recreate the third pic, but it doesn’t, it recreates the second pic again. So, something about interpolation carries over in those sensors even if I declare them bad again. I tried this with the bad channels as the original bad data and again with zeroed out channels; both give the same result.
Any idea what is happening here with the interpolation? It is not a big change, but it is a mildly concerning one. I am not sure if I should commit to these changes, or just try re-preprocessing them.
Best,
Matt