Val’s Views – On the Road With a MaverickZengar Institute Inc.
“In 1970 I had a vision of what brain training could be in the future: a simple, powerful, and safe way to self-optimize that is so easy and intuitive that anyone can use it. NeurOptimal® version 3 is that elegant vision and it’s here now.”
Why am I starting with this? Because that vision has animated all of my work as well as the work my wife Susan Brown, Ph.D. and I did to co-develop NeurOptimal®. That work has now been determined by the FDA to involve a General Wellness Product and that means anyone can use it: no special training, knowledge, expertise, certification of licensure required or needed. This orientation has been one of the most distinct contributions I have made and it brought enormous animosity from many others in the field of neurofeedback. So what went into developing “Dynamical Neurofeedback® and what makes it so different from every other approach to neurofeedback? My specific contributions did couple with Dr. Susan Brown’s work with the evolving system.
In 1970 I had just finished reading Freud’s Project for a Scientific Psychology <ref from Std Ed..> (also known as Psychology for Neurologists) and I realized that energy transformation was the core of the work. The problem was that Freud’s metaphor was mechanical, viz the steam engine. So I thought that, if we substituted the energy transformations of quantum mechanics, it would be possible to implement a technology that could help support personal transformation using computers to handle the calculations needed. Quickly extrapolating how much computing resource would be needed for that, I understood that it would likely not be until some time around the year 2000 before it could be done economically. So I put the project on the shelf.
In 1972 I had the good fortune to hear Karl Pribram lecture on the work that he was doing with Merton Gill to interpret the Project for a Scientific Psychology in more modern terms, in line with the Holonomic Model of Consciousness, Memory and Perception that he had just published <ref>. I asked several questions about the specific mathematics he used for that work and I then knew that my original vision was possible…in time. Again the neurofeedback vision I had was put back on the shelf. In 2000 I presented at the FutureHealth Conference on the work that Sue and I were doing to develop our own neurofeedback system. Karl was there, approached me during a break and said that what I was doing “…summed up the last 25 years of <his> work.” I thanked him and said: “You probably don’t remember me but I heard you at Georgetown…” He looked at me more closely and asked: “Were you that long haired young man in the back of the room that asked about the mathematics?” I said, yes, and that began a very important relationship that quickly led to Karl and me doing joint presentations at conferences on the Holonomic Model of the Brain and how to approach neurofeedback with that in mind as I was doing.
The mathematics involved were specific modifications to Joint Time-Frequency Analysis techniques used by their developer, Denis Gabor, to develop Holography: an achievement that earned him a Nobel Prize in 1972 <ref>. When I brought the term Joint Time-Frequency Analysis (JTFA) into the field of neurofeedback in 1998, many commented: “I’ve never heard of that”. I found that astounding because no matter what else you do in neurofeedback, if you are concerned about how energy flows through time during a Session, you are doing JTFA. The only question is how precise, sophisticated and useful are the ways in which you do consider frequency and time and how they interrelate within neurofeedback. Why was the proper JTFA technique important? There are two reasons: 1. they allow you to do inline adaptive denoising of non-human EEG components of the signal recorded from the scalp and 2. these are the same mathematics that the CNS itself uses to perceive, remember and navigate through the world, as Pribram had demonstrated in the Holonomic Model he developed <ref>.
How can the right JTFA denoise EEG, removing as much as possible artifacts like Line Noise, EMG, physical movement, etc? Because those artifacts have different time-frequency event structures than EEG. Line Noise, for instance, is essentially a noisy, simple sinusoid so it basically is repetitious, predictable and always maintains the same basic frequency. Human EEG never does this unless in the grips of a grand mal seizure <ref>. JTFA-based targeting (similar to what is used in Doppler RADAR, etc) looks for specific time-frequency events, and this makes the right use of sophisticated JTFA ideal for providing neurofeedback, esp in the midst of non-human EEG artifacts. The mathematics are elegant, clear and precise but require enormous amounts of computing power and that was the challenge: viz, how to bring sufficient computing power to a system that could be cost effective? It was a question of Moore’s Law and by 1998, we were on the cusp of having consumer-based, cost effective personal computers that could begin to support this specific form of JTFA.
So now the question became: how can this process be automated so no human operator is needed, ie that anyone could use the system at any time, with anybody, including themselves? An algorithm needed to be developed that could automatically tune the JTFA-based targeting to the emergent the specific dynamical signal during each Session, and that algorithm had to be able to individualize that emergent tuning to each separate person who used it within each distinct moment of each Session. This meant jettisoning any form of normative databased approach.
Again a bit of my personal history becomes relevant. I had the good fortune to begin to explore the emerging mathematics of Nonlinear Dynamical Systems (NDS) while in High School and that provided the basis for the above described algorithm that eventually became AutoNav. This sophisticated and unique algorithm allowed the computer to tune the specific textures of each moment in each neurofeedback Session based on how that particular Client responded to the prior moment in that specific Session. The method is not based on averages, medians, standard deviations, standard approaches to variance, etc but on specific NDS techniques further refined to the unique demands of neurofeedback. When AutoNav was implemented in combination with our unique, proprietary JTAF-based targeting approach, we had a Comprehensive, Adaptive method of neurofeedback that could respond optimally to each person as they were using it: no prior assessment needed, no change in protocol, targeting, or sensor site needed!
Another contribution is the use of Interrupts in ongoing audiovisual streams instead of discrete positive feedback stimuli, commonly referred to as “rewards”. Instead NeurOptimal® sends an Interrupt to its Media Player and the ongoing audiovisual stream has a very brief pause which can sound like static, or the clicks and skips that we used to hear when playing back vinyl records. This allows for each Client to use whatever audiovisual resource they enjoy, although most use the media that we supply with the system. Rather than being annoying or unpleasant in any way, the Interrupts are basically not processed consciously; rather, they work like the rumble strips on highways. Those just give you information: if you want to pull off the road keep going, if you want to remain on the road, turn back toward your lane. No need to count the number of bumps to figure out if you’re doing it right! Similarly, there is no preferred amount of Interrupts to have, in fact there are no specific tasks to do, or goals to aim for when using NeurOptimal®. Instead we give information to the CNS through the Interrupts and that CNS itself chooses what to do, utilizing its own intrinsic negative feedback processes to maintain its resilience and flexibility. Virtually all other forms of neurofeedback are based on some form of positive feedback and that carries with it the inherent risk of overshoot. In the early day we used to use different, singular feedback protocols to promote certain states: one would relax you whereas another would increase your activation. Back then you knew you had to switch protocols because you could easily overshoot the intended activation and need to bring that back down through relaxation. There is no such phenomenon with NeurOptimal®: its AutoNav algorithm smoothly tunes the comprehensive, adaptive array of our ten pairs of bilaterally symmetric Time-Frequency Envelopes (or Targets).
Another contribution was letting go of “state based” ideas and looking instead into how the CNS navigates its world. I originally did this in recording Baseline recordings of 30 seconds: the first 15 seconds the Client’s eyes were open and looking at a computer monitor, and the last 15 seconds the Client had eyes closed. I was looking at how the CNS navigating that transformation and when I presented that idea at conferences the remark was: “We already know the difference between eyes open and eyes closed EEG”. I replied I was investigating the flow or transition itself, not separate, presumed “states”. I was astonished that it was difficult for others to hear that difference. Letting go of “states” and state-based training makes perfect sense when one realizes that the CNS is a nonlinear dynamical system and those are never in just a single “state”; rather, there are always multiple overlapping and interpenetrating processes that arise and then fall away in a syncopated dance of navigating the world. How can there be a single “state” unless you “pin the butterfly to the wall to see how it flies”?
Overall I believe these contributions are unique and noteworthy. I have not referred to any other contributors or their contributions because I know that others here will.
– By Valdeane W. Brown, Ph.D