How to rule out low SNR's influence over functional connectivity?

Hi Mike,

In your ANTS book, chapter 19.6 and 26.5 talk about ITPC/ISPC and power. It was proved that connectivity would not be influenced when the signal amplitude was dampened or enlarged considerably. Therefore, connectivity could be interpreted independent of power as long as power is not zero. However, when the amplitude of signal was adjusted by multiplying some extreme numbers (e.g. 0.0001), the signal-noise-ratio (SNR) would not change. Just as the paper Muthukumaraswamy and Singh 2011 showed, low SNR affected connectivity. My question is how to rule out low SNR’s influence on functional connectivity (e.g. ISPC, WPLI).

All the best,

Hi Jinwen. Good question. But “SNR” has multiple interpretations, so you’d first need to pick a definition for SNR (I also discuss this in the book somewhere).

After that, I’d recommend going through the steps to make sure your connectivity results are not due to artifacts of volume conduction, where I provide a few suggestions to disambiguate power (which is sometimes taken as a measure of SNR) from connectivity.

Thanks Mike. I’ll try it.