[Eeglablist] average for frequency power scalp maps

[Cedric Cannard]

Hi Ana,

It depends on your aims.

If you want to give it a try, see this repo/plugin: 
https://urldefense.com/v3/__https://github.com/amisepa/reference_infinity__;!!Mih3wA!EDxhk3mqeJCy_UGKTQqiArOMFn4SafON3YFzszkYVUPrAaN00e9qjXTCShuPXJs9lpZpl8qvxFqu2XFMWdzvyg2JOQ$ 

Or you can also perform current-source density (CSD; surface Laplacian) transformation with this one: https://urldefense.com/v3/__https://github.com/amisepa/csd_transfrom__;!!Mih3wA!EDxhk3mqeJCy_UGKTQqiArOMFn4SafON3YFzszkYVUPrAaN00e9qjXTCShuPXJs9lpZpl8qvxFqu2XFMWdyhaO2XyQ$ 

These referencing/transformations aim to address volume conduction and reduce the mixing of signals as they pass through the skull, while still operating at the scalp channel level. The goal is to improve spatial resolution by emphasizing local dynamics and reducing the inflated global effects typically caused by volume conduction. Most experts see methods like CSD or Laplacian as useful tools that sharpen scalp topographies and highlight local activity, but they also agree that these are not perfect solutions. Their effectiveness depends on assumptions about the head model, which are rarely accurate for each individual, so results should be interpreted with caution.

At the same time, common average referencing (CAR) is also imperfect. It mathematically assumes uniform spherical coverage, but this is never achieved in practice since electrodes do not cover the neck (or face most of the time). And as noted in Nunez and Srinivasan in Electric Fields of the Brain (2006, 2nd edition), at least 128 electrodes are needed for the approximation to be reasonably accurate (rule-of-thumb which is largely not respected by most). 

Makoto will probably provide a better assessment on this. 

Cedric Cannard



On Wednesday, September 17th, 2025 at 9:48 AM, Ana Navarro Cebrian via eeglablist <eeglablist at sccn.ucsd.edu> wrote:

> Dear all,
> 
> I have been using the average as my reference for years. I am now really
> interested in getting the most reliable frequency power scalp maps and
> would like to know what people's thoughts are about the best reference to
> use for this. I was just looking at Yao, Wang, Oostenveld et al., 2005,
> paper and it has made me doubt about the use of the average for power maps.
> I may also end up looking at frequency correlations between channels.
> 
> Thank you in advance for any advice.
> Best,
> Ana
> _______________________________________________
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu or visit https://sccn.ucsd.edu/mailman/listinfo/eeglablist .