November 9, 2003

Report on ISNR

Author: Siegfried Othmer

I left the ISNR Conference with my head full of rich detail, but over time it is easier to discern the deeper trends. First of all, the conference was anchored by some top mainstream scientists who have seriously bitten into neurofeedback. They are now very committed to this field, and they will carry on through their publications and graduate students, and through presentations at other conferences. They will be bringing up the rear with sound research. There were also first-rate presentations by the indigenous leaders of our field. And new capabilities are being developed on the instrumentation side. Quite unambiguously the forefront of this field is still in the hands of the scientist/practitioners, the clinicians, and the engineers.

On the matter of stimulation
Perhaps the development of greatest import at this time is the maturation that we see in the area of the stimulation technologies. I have been watching that development with intense interest over the years, but there was always a serious philosophical barrier. There was this conceptual divide between active training and passive driving, and we have always preferred to stay on the training side of that divide. This calls upon the active involvement of the client or patient in the work being done. It involves a learning component. We can say unabashedly that we are not doing anything to the brain. This is all very comforting, both to the clinician and the client, and it keeps the technique more unambiguously in the domain of psychology.

Two things have changed. The first is the recognition on my part that the greater progress in neurofeedback efficacy is to be made near-term by the judicious handling of the inhibits than by increasing refinement of the rewards. Certainly in our own part of the neurofeedback world we have been focusing largely on the rewards up to now, and the inhibit strategies remained relatively less developed. Val Brown made a conscious shift in the other direction with his multiple targeting strategy, and there has already been a significant payoff for that. But the fact remains that the inhibit function is still largely indirect: we simply withhold rewards. This gives the brain no subtle cues to altering its behavior. It is too coarse a remedy. Something else is needed.

The other change that has occurred is that we have increasingly come to diminish the distinctions that in fact prevail between the training and driving model. We have thought for some time that there is really a two-stage process going on here. The brain is coaxed out of its prevailing state, and then it is moved to restore something like the original state. That is, the brain is not tolerant of arbitrary displacement out of its bio-electrical state. It will react, and out of the repetition of that push-pull process eventually comes learning. It could be argued that it is of somewhat secondary importance how the original displacement was effected, whether by operant conditioning or by stimulation. What is quite clear by now is that learning occurs with either approach, and lasting changes in brain behavior may be produced either way.

It is also true that the consent being given to this process by the client is just as “provisional” or “contingent” either way. A person cannot really be said to have given consent when the issues are not understood, and when the choices are not clear. The consent is also a function of time. It grows as the migraine goes away, so to speak.

The client has given a general consent to have the clinician “train his brain,” and he has given specific consent to try to make Pacman go. The intermediate step, namely that making Pacman go increases the beta amplitude at C3, cannot really be appreciated for its import. So that particular aspect of consent cannot be very meaningful, and we don’t dwell on it. If the EEG amplitude is instead transiently altered by other means, but in the service of the same objective, has the consent of the client been significantly violated? Before that question is answered, we should cover a couple of other considerations:

The stimulation technologies that are just now being brought to maturity within our field are responsive to the EEG. They are driven by the EEG. This differs substantially from the prior art, in which Light and Sound systems, or Audio-Visual Stimulation units (AVS), delivered a tailored or programmed waveform. Such EEG-responsive signals are perhaps better regarded as specific cues to the brain, i.e. individualized feedback much as we do with the rewards. Only in this instance, the feedback is to the brain more directly. It is obviously more direct in the case of ROSHI’s MagStim and Len Ochs’ radio-frequency imbedded signal in his LENS system. But it is also more direct with the visual signal (the LEDs), because information is contained in the actual waveform, not merely in the abstraction of the size of an object on the screen. We are dependent simply on reception of an optical signal by the brain, and not on the perception and interpretation of a visual image.

In other words, we would not be far off the mark in regarding this as a feedback signal rather than from the perspective of stimulation. This conception is reinforced when we recognize that what is being accomplished here is disentrainment of existing EEG patterns, rather than the entrainment of patterns that we associate with stimulation technologies. This is really the first active inhibition process available to us. One may still raise the concern that we have now breached the barrier of non-invasiveness of our methods. But let us take a closer look. What is usually meant by invasive technologies is something that addresses issues of structure rather than of function. And in that regard, we are clearly on the side of function. As to invasiveness, the modulated light could be said to function much like binaural beats, which we would not regard as invasive. And the modulated magnetic field functions much like the modulated light. So we must resort to creeping moral relativism.

In this manner, we incrementally rationalize the adoption of this kind of direct, instantaneous, non-volitional feedback. Readers will know by now that in recommending this we are going very much against type. But we cannot simply fail to take advantage of the benefits that have been demonstrated with this technique to our most challenging populations. The issue remains just how we are going to explain this to clients.

Just how do we handle the signal now in our discussions with the client? We insist that the person not try to figure this process out on the conscious level, but rather to allow the brain simply to process the signal and to engage with it. Consciousness hangs somewhere between a hindrance and a help. Already we recognize that consciousness plays a very limited role in neurofeedback, and none at all down in the trenches. So why not bypass consciousness when the opportunity allows, provided the person has given the real “consent to treat” that matters?

There is at least one reason to consider. People derive a sense of empowerment from this work, and whereas much of that comes from the functional improvement we achieve, there is also the subjective experience of doing the work yourself. This subjective component is probably intact even if only part of the feedback involves volition. So as long as we meld the volitional and non-volitional constituents of the feedback together into one treatment, we should be fine on this count.

So we have persuaded ourselves that when the reward strategy via conventional neurofeedback and the EEG-driven inhibit strategy via visual or electromagnetic stimulation are combined into one overall treatment approach, it passes muster to think of this entirely in terms of feedback. At the last CIC, I made the case for improving the inhibit strategy by working on shorter timescales and with more subtle signals, that by doing this we could match the training “efficiency” that now attaches to our rewards also with the inhibits. One can approach these objectives within the traditional feedback paradigm (and we’re moving along on that front as well). But nothing accomplishes this as well as the devices devised by Chuck Davis and Len Ochs. The feedback signal tracks the EEG, and this involves minimal delays in the signal processing chain. To be effective, this kind of feedback simply could not be provided in the conventional way. What remained was to show that the brain actually learns from that kind of cueing, and that issue was favorably resolved for both technologies quite some time ago.

So at this conference both Chuck Davis and Len Ochs announced their second-generation systems. Chuck’s system was on exhibit, and the booth was busy throughout the conference. Chuck’s first-generation system depended on an Amiga computer, which lately have been harder to get. For a while they were still manufactured in Europe. The new system interfaces with the BioExplorer software for EEG display. Retail price is $2995 with the BioExplorer software, $2495 without.

We are moving into a more multi-polar world in neurofeedback. Whereas earlier clinicians might simply have added additional units of their standard neurofeedback system as their practice grew, nowadays it is more likely that they will gather multiple technologies under their roof. Within that philosophy, the ROSHI would be a worthy addition. The ROSHI does not yet have FDA exemption, however. There are two more points I need to bring up in the spirit of full disclosure. Chuck has promised to give a unit to the EEG Institute for our usage and evaluation. For that we are very grateful. And Kurt has a dealer relationship with Chuck through EEG Support.

How does Len Ochs’ system fit in here? It is based on J&J’s electronics, which is an excellent platform. The system differs from Chuck’s in that the feedback or stimulation signal is delivered right back to the electrode site from which the EEG measurement is derived. So the feedback is more targeted. The system is also more integrated with a training strategy in which various scalp sites are successively trained, based on the results of a minimap that shows how responsive each site is to the stimulation signal. Training migrates sequentially from the most robust or stable sites all the way to the least stable. I am content to be brief here for two reasons. The first is that this system is also described in an earlier newsletter that can be found on the Bulletin Board, and the second is that Steve Larsen has offered to write a full newsletter on this technology from the perspective of an enthusiastic user. That’s worth a lot more than my speculations. More information can also be obtained from the website, www.ochslabs.com

With the shift in perspective to the inhibits for a while, we are seeing another of the inevitable ratchetings away from overt behavior to abstract brain behavior. In our mechanisms-based models we still tie our conceptions as much as possible to behavioral considerations. In the QEEG world, too, the assumption is rashly made that whatever deviant pattern is seen is explained by the diagnosis the person came in with. (We are such pushovers when it comes to data!) Val Brown has already cut the umbilical between what he does with his technique and what the person complains of, although probably more in rhetoric than in practice. The trend will continue. This will increasingly become generic brain training, unmoored from diagnosis. The inhibit function is a disregulation detector, and the target is therefore disregulation itself---agnostic with respect to clinical categories or specific symptoms.

There has been some discussion lately about whether the burden of training can be largely borne by the inhibits. Margaret Ayers says that she has shifted her perspective in that direction. And Peter van Deusen uses even the inter-hemispheric derivations with inhibit-only training to good effect. Are we so sure that the reward is actually essential to what we do? We are sure that it changes the person’s state, and that has consequences both good and ill. But how much is that process actually contributing to the remediation? The stimulation technologies provide an answer. At the ISNR, Dave Siever presented a workshop on the use of AVS. I was unable to attend, but the already published work by Michael Joyce and others makes clear that AVS can be quite helpful with ADHD, etc., with the normalization of TOVAs, etc. And significantly there is no inhibit component at all associated with this technique. It is all work at a single frequency at a time. And of course Barry’s original work with the cats stands as a landmark for targeted mechanisms-based training. Jointly, this makes the case, if any were still needed, that the reward plays a crucial role for us as well. It is indeed worth all the trouble.

There is now one more contender in the area of stimulation technologies, and it is rTMS, which stands for repetitive Transcranial Magnetic Stimulation. I was able to attend the workshop given by Steven Padgitt on this topic on the last day of the conference. Steve has been using the Solomon II, an instrument developed in Korea, in his practice for some time. He finds it to cover much of the same ground as we do with neurofeedback: depression, migraines, etc.

In contrast to the ROSHI and LENS, the Solomon II is an entrainment rather than disentrainment device, by analogy with AVS. What distinguishes the Solomon II from the clinical technologies in use in this country is that it uses a much lower level of stimulation, and it does so for longer durations. The American units that are under investigation as part of NIH-sponsored studies use very large magnetic fields—up to 20,000 gauss. The Solomon II, on the other hand, uses 135 to 500 milligauss. The upper end of this range is about the same magnitude as the earth’s magnetic field, with the difference that this field is modulated. And that of course makes all the difference. The EEG readily entrains to the periodic activity even at this low level.

Because the levels are so low, the FDA has allowed the unit to be sold in this country for self-improvement purposes. The levels are in fact quite adequate to accomplish any of our purposes. Steve has seen salutary results for conditions ranging from ADHD to Bipolar Disorder. Right now the instrument is programmed with a number of standard routines for different purposes, but one can readily imagine an even greater repertoire tailored to different applications. The simplicity of the device and the canned nature of the programs also lends itself to home use. One can imagine speeding up the pace of neurofeedback by having people use the Solomon II at home a few times between visits to the office. One can also imagine a bipolar person having one available, or a person at risk of suicide, or a person who is at risk of a lapse back into drug use, or a migraineur at the beginning stages of a migraine, or an insomniac.

The advantage of the Solomon II over the AVS technologies is that the stimulation is localized. It has this in common with Len Ochs’ system, but the latter, being EEG-responsive, is of course much more involved. Whereas one would never dream of putting Len Ochs’ system into home use with someone, the Solomon II gives rise to no such concerns. A dual coil unit has also been developed that can be used at the common homologous sites such as C3 & C4, for example in application to migraines, and perhaps to seizures as well.

One can also imagine using the Solomon II coils as quick probes of how someone is going to respond to particular placements. The unit is relative inexpensive, on the order of $500. I would love to see some experimentation by some of the network practitioners to see what it can offer us in complementing our neurofeedback strategies, particularly for home use. (Contact Steve Padgitt at: http://www.brainwavecenter.com/products/solomon/next_step.htm)

Another overall trend that is discernible is the proliferation of technologies and of protocols within a technology. We have as much as anyone attempted to keep the training strategy simple, and Val Brown has embedded a lot of complexity in an approach which was to offer simplicity to the user. But it was not to be. We are now all over the scalp with our training. And Val’s software offers many alternative “deployments.” I’m convinced that a client would not have the same training experience with Val as he would with his wife Sue Dermit Brown, even though the equipment is the same and they have shared perspectives. It’s not a cookie cutter approach. Then there is pIR HEG and nIR HEG, and now there is ROSHI, LENS, and rTMS. Each of these has its uniqueness.

So, what’s the message out of all this? First of all, I think we as a professional community need to embrace the multi-polar world that is developing. The experience at ISNR this year shows that is happening. Many of you will be content to just hang on with your NeuroCybernetics unit or BrainMaster, Biograph, BioIntegrator or other system you are using, but others will not want to leave major advances on the table. The ROSHI is probably the next Big Thing, advantaged as it is by elegance of concept, simplicity in usage, compatibility with neurofeedback, and now finally an interface to a Windows system through the BioExplorer. The LENS system is the top end system for this kind of work, but it will have somewhat narrower appeal because it is a bit more daunting. Those of you who are ready for that kind of challenge know who you are.

I have dwelt mainly on the non-volitional neurofeedback options in this newsletter. I will cover other issues in newsletters to follow