Sum Difference (Sum Diff)
Sum Difference (Sum Diff)
In simplest terms, I use Sum Difference when there is a strong differential in a group of frequencies between homologous sites and there are one or more frequency bands that show excessive activation. Training down the sum trains down the activity in both channels, training down difference reduces the asymmetry. You can also use a bipolar montage to achieve the same effect.
The sum difference designs allow you to choose the bands that are most asymmetrical for your difference training and those which are most out of line with the rest of the brain for the sum training. You can use any combination of the options.
The Sum Difference protocol trains down the total activity at two sites in one frequency band and trains down the difference between the two sites at another (could be the same or different band at the same time. A person with a lot of activity in the very fast (19-38 Hz) band who also has much more on the right side than on the left could train down the sum of the two (reducing overall fast activity) and, at the same time, train down the difference in that frequency between right and left, making them more alike.
I produced this approach as an option to 1C bipolar training. As has often been mentioned on the list, bipolar training has the potential problem that as we train down T3-T4 we have no idea whether the brain is responding by reducing the side that is too high or training up the side that is low. By training in two channels, you can see what is happening. And it is argued that training down T3-T4 doesn’t necessarily reduce the total of the two, whereas Sum Difference is challenging the brain to do that as well and giving us information as to whether that is happening or not.
When your assessment shows two sites that have a large asymmetry in one or the other in specific frequencies, for example, a lot more high beta in CH2 than in CH1, you could train to reduce the DIFFERENCE between the two sites. Then you would look at the histograms page and see if there is a frequency band (for example 2-6 Hz) where both sides are very high. You could train to reduce the SUM of the two sites in that frequency.