What the EEG Captures

The EEG is a summation of positive and negative signals at one point subtracted from the summation of positive and negative signals at another point–NOT just the number at a single point.  What’s more, Tom Collura and Val Brown did an excellent presentation at WinterBrain several years ago, demonstrating that the EEG we measure is primarily measured from pyramidal neurons only (they have a positive and negative end, so there is a difference in the reading at the two electrodes), and primarily measured from those that are lined up parallel to the line of measurement.  If you are reading between C3 and C4, for example, neurons at Cz which are oriented across the measurement line (running, say, between Fz and Pz) will be completely invisible to the EEG.  Since the cortex is a folded and crumpled sheet of columns of neurons, individual neurons and even pools of them are oriented in many different directions, so the EEG measures only a fairly small sample of those that are in the area we are measuring from.

The only places we can see on an EEG are areas which have pyramidal neurons instead of stellate cells. That means that we can see the cortex, the hippocampus and the cingulate. The rest is invisible (though not to magnetic readings). I don’t necessarily think that means we can’t train other areas. For example, training up peak alpha frequency certainly has an effect on the thalamus, as does training SMR in the central strip.

The EEG also is made up of a lot of background noise from neurons far from the measurement sites (which is why most EEGs are fairly homogeneous).  Since the brain is a volume conductor, a saline-soaked sponge trapped inside a very poor conductor, any signal anywhere in the brain can be seen most anywhere else in the brain.  It will just be a lot weaker.

The bottom line is that looking at what a single neuron does can no more directly imply what we will see in a measurement of differences between two sites than looking at how one person in New York City lives can help us predict the differences we will find between New Yorkers and Atlantans.

Thus, one of the rules of neurofeedback is that if you train anywhere, you probably have an effect everywhere.  Much of the signal we see in the cortex is noise, actually firing patterns from neurons all over the brain.