Hi choepel, could you explain what CCAC is? I understand it is "cross correlation of auto correlation". What is "auto correlation", specifically what is correlated?
Technically, this is complicated and would require an excellent understanding of mathematics to understand. If you are one of those people (I am not) I have posted two passages from Val below which may make sense to you. The particular mathematical calculation is unique to NO - it is VAL's invention as I understand it. It is calculated from the baseline spectrograms and depends on the client closing their eyes half way through, as this is the "phase shift" that is mainly being measured. Cross correlation and autocorrelation are specific signal terms which can be looked up on e.g. wikipedia.
In essence, the CCAC is an attempt to measure the resilience and flexibility of the brain and to measure the impact the sessions are having on the brain. The smoother the curve and the lower the divergence number, the more resilient the brain. My most psychologically injured client (a lady who had 10 years of Childhood Sexual Abuse and later rapes) had the best curve I have ever seen. But she was resilient and despite her alcoholism and self-destructive behaviour, she was the editor of a major newspaper.
But VAL warns against putting too much emphasis on the CCAC - in fact in recent years he advised to stop using baselines because he said they could be very misleading and could be easily misinterpreted. Instead, he emphasised using the client's own experience of change. In NO3 he has replaced CCACs with tunnels which are a similar attempt but using more data and being somewhat non-linear in their analysis.
How is the CCAC calculated (Val Quote from Webinar)?
Note: Joint Time Frequency Analysis is at the heart of NO - it is a type of fourier transform)
Taking the joint time frequency analysis of the journey, using the spectrogram data auto correlating each of those channels to themselves multiple times, and the cross correlating each of those auto correlations to derive a single divergence curve.
It has nothing to do with relative power, volume or amplitude or even special distribution in the brain.
It’s a measure of efficiency and effectiveness or resilience and flexibility
The smoother the curve and the lower the divergence (as long as it is not too low)
Over sessions, the divergence difference number should decrease. But that presupposes that the sessions have similar levels of line noise and conditions – similar amounts of paste being used etc.
Val Quote:
Essentially divergence is an indirect measure of "wobble" in the overall system. It is a measure that I developed based on the Cross Ambiguity Function used in Joint Time-Frequency Analysis (JTFA). I say that it's "indirect" because the absolute number that it produces depend to a degree on the relative amount of Line Noise present in the Baseline you are Analyzing. For this and several other reasons it doesn't make a lot of sense to talk about the meaning of a single divergence number; instead it is more useful to look at the divergence difference between a pair of Pre and Post Baselines from the same session. This give a measure of relative change in flexibility and resilience in the CNS as a result of Training. The more resilient and flexible the CNS (or in general any other nonlinear dynamic system) is, the higher the divergence will be. However, extrinsic constraints (yes this is a term to also search) like Line Noise can significantly alter the divergence "number" even though that doesn't affect the value of the Training. This is because of our unique array of filtering processes that essentially remove any simple sinusoidal signals like Line Noise. Regardless of what others may have said, human EEG is not composed of simple sinusoids, and trying to use mathematical approaches based on Short Time Fourier Transformations (STFT) are fundamentally deceiving. This applies to any kind of QEEG or similar databased approaches to determining what others call an "individualized" neurofeedback approach. NO does this kind of resizing dynamically, moment-to-moment based on the actual Training response of that particular Client at that time; not based on some "snapshot" taken before the Training -- or even after it!
