How do I "normalize" the wave functions of each Channel based on their impedance?

lagidigu

Hi there!

I am currently making a little software for microstate analysis. For that, the signals need to be "normalized" based on their impedance,  so that the signal strength doesn't vary from one recording session to the other just because of a change in contact quality. Does anyone know any pointers on how this is computable?

Thanks a lot in advance!

Luc

wjcroft

Luc, hi.

EGI (the manufacturer of the Geodesic Sensor Net silver chloride saline system) documented that even skin impedances in the range of ~50K ohms or higher had no impact on the measured EEG. Compared to older EEG amps that required ~5K ohm impedance. This is with high impedance amps such as the ADS1299 used in Cyton.

http://openbci.com/forum/index.php?p=/discussion/165/measuring-electrode-impedance

William

lagidigu

Thanks for the quick answer!

That would seem quite trivial then!

Does that mean that the difference between the same measurement done with 5kOhms and 10kOhms is negligeable, or is there simply no difference at all? 

My main worry is that the signal tends to be much stronger on certain sites on my head than others, while each channel has a maximal impedance of 10kOhms. So if I understand you correctly, this isn't due to the impedance per se, but due to the potentials generated on the sites themselves?

Cheers

Luc

wjcroft

Correct, difference between 5k and 10k will be negligible.

The sites that have higher amplitude are not due lower impedance, provided all your electrode impedances are in the same ballpark range.

Because the Cyton amp has such a high input impedance (1 gigaohm), small skin impedance differences should not affect your readings. It could be that the microstate normalization (in a research paper?) you refer to, was done on an older amp with impedance in say the megohm range, which IS more sensitive to skin impedance.

lagidigu

That is awesome to hear!

Thanks a lot William!