Inhibition and Reward

by Siegfried Othmer | September 8th, 2005

With all of the disagreements outstanding in this field, the least we can do is clear the semantic hurdles that may stand in the way of synthesis. One issue that is still outstanding, but can easily be dispatched, concerns how we think about inhibits. In the past I have variously contrasted Sue’s largely reward-based training with Val Brown’s largely inhibit-based approach. The distinction is quite clear to me in my own mind, but Val Brown comes back to point out that his system does allow the promotion of amplitudes in particular bands as well. In his approach of box targeting there is both an upper and a lower threshold set for every band. One can choose to raise the lower threshold with an imposed offset, thus rewarding the brain for larger amplitudes in that particular band. That could then be seen as a reward strategy. (The upper threshold gets raised as well, in that the whole box is moved upward, but that is not of interest at the moment.)

Now in fairness it must be said that Val has moved away from the use of this terminology of augments and inhibits in his own chosen framing of what he does. So the problem is not his but ours, as we try to understand his approach in our traditional framework. On the other hand, Val also has an interest in not having his system misrepresented. So, why do I still talk of Val’s approach in terms of inhibit-based training?

As George von Hilsheimer pointed out some long time ago, our little enterprise in no wise inhibits the brain. What we actually do is inhibit the feedback signal on the screen, thus provoking the brain into some kind of response. Taking that thought forward, any cessation of feedback on the screen is in our world an inhibit strategy. And when Val uses that strategy to promote larger amplitudes in a particular band, then this encouragement of an augmentation is still being done with an inhibit strategy.

As is the case with inhibit strategies generally, they must be applied sparingly in NeuroCarePro. No child is going to sit still for looking at a mostly blank screen. Thresholds indicate boundary transgressions, and as such they must indeed be rare in order to be effective.

Matters are totally different when it comes to the rewards, as these have evolved over the past two decades. The essential difference here is that the reward strategies reflect back to the brain its ongoing dynamical functioning. They give the brain an opportunity to observe itself in action, and to do pattern-matching between what is going on “out there” with what is going on inside the brain. No other explanation can account for the rapidity of response to the particulars of reinforcement.

It is this immediacy, this continuing representation of the brain back to itself, that has been relatively absent historically in NeuroCarePro. The dynamics displayed on the screen in NCP are typically unrelated to what is happening in the brain. They are mere entertainment. It is for this reason that the bulk of what NCP accomplishes must be seen in terms of the inhibit strategy, i.e. through the occasional disruption of the visual and auditory reward. Now things might well have changed with the recent evolution of NCP, but Val is keeping the particulars to himself, so I cannot comment.

It is therefore of particular interest to me that so much can be accomplished with what I consider to be largely an inhibit-based strategy, and that there is so much overlap with what we are accomplishing with a largely reward-based strategy. Firstly, this implies that there are many pathways to the goal of improved self-regulation. Secondly, it suggests a merger of the two strategies. Just as the NCP strategy has historically been weak on the reward-side, the single-channel inter-hemispheric training has been relatively weak on the inhibit side.

The combination of features of both is of course what we hope to accomplish with our imminent transition to two-channel training. Here the opportunity is presented to inhibit on the emergence of excess synchrony between channels along with the promotion of desynchronization with the continuous reward-based strategy.

There is an interesting reversal of roles here. Val has emphasized brain dynamics in the discussions of his model, yet brain dynamics require significant signal bandwidth for their representation. On the face of it, a dynamical model of neurofeedback would seem to be much better represented by our continuous reward-based training than by the much lower signal bandwidth of the inhibits. After all, we too have moved away from the concept of augments that we started out with on the basis of Sterman and Lubar. We may still be rewarding the person for larger amplitudes instantaneously, but we are not necessarily aiming for larger amplitudes as the end result. The brain challenge serves as a stimulus toward engagement with the information, and a richer information flow only facilitates this process. We really are concerned with the evolution of brain dynamics in real time, and arranging for the brain to be a witness. Goals in the training may give the whole thing more drama, but the more important aspect may be the facilitation of engagement.

By contrast, the inhibit-based training can be seen by analogy to a random walk with hard walls. The encounters with the walls are rare. Hence the signal bandwidth is low. Hence brain dynamics cannot be well represented there. The brain is clobbered every once in a while, and it is moved to avoid a reoccurrence. A brain-dynamical model may underlie this strategy, but the brain remains largely uninformed of the particulars.

Given the fact that our kind of reward strategy does clearly operate successfully with a signal bandwidth of nominally three Hz, the burden would seem to fall on anyone who is inclined to claim that such larger signal bandwidth is irrelevant. We not only find it relevant, but it underpins everything that we do. I therefore believe that we should not settle for any approach that leaves such higher dynamics out of the picture.

Fudgie Lives!

It may not be a good idea to get one’s science news out of the newspaper, but the press is often a good source of anecdote. Dennis McCarthy retells the story in the Los Angeles Times of the adventure of Fudgie, a pet dwarf steer that had eaten some poisonous oleander leaves in such quantity that he should have died twice over. The story previously appeared in the Journal of the American Veterinarian Medical Association, which is not on everyone’s reading list. Fudgie was treated by one of the few vets that still deals with cattle around these parts. Fudgie was barely able to stand up, and sometimes lost consciousness. In time his heart stopped and he collapsed to the ground. The vet kicked him each time to get the heart restarted. Every few hours the process would repeat. The EEG would flat-line, the eyes would roll back in his head, and he would collapse. A kick would get things restarted.

The vet sent the electrocardiogram over to a cardiologist, and word came back that Fudgie had every heart arrhythmia known to man. Drugs were tried as well, but there is no history of success in such treatments, so there was little to go on. “I’m not sure we’ll ever find out why he lived,” said the vet, “but by following the medications given, it will help with future cases…” The vet and his staff had put the steer on twenty-four hour watch in one-hour shifts to keep him alive. The people there had kind of adopted him.

It seems obvious that what kept the steer alive was the kicks delivered to the heart whenever that was needed. Whatever the medications did would not have been sufficient. The heart rhythms were at times unable to maintain their organization, and the heart would cease to function. A well-delivered blow would resynchronize things for a while, and restart a heart that was really trying to continue to function. It was all about timing. It is an experiment better done first in animals. But surely there is every reason to think that these results would apply to man as well.

So, with reference to our other themes of interest in this newsletter, is kicking the patient to be regarded as a medical procedure? In the context of the overall treatment of the patient for his heart condition, and in the hands of a medical professional, the answer is yes. But in the context of an emergency in which someone’s heart has stopped where medical care is not immediately available, this “procedure” would be available for anyone. Which brings to mind what we might call “The Good Samaritan Model.”

The Good Samaritan Model

In an emergency a person may give aid to another, and that would even include some medical procedures. This may be the best that can be done under the immediate circumstances, and the world is forgiving even if the attempt were to be a clumsy one. This is codified in “Good Samaritan” laws. Can this principle not be extended to situations in which care is not available for other reasons, even outside emergency situations?

Suppose, for example, that neurofeedback were to be declared a medical procedure that could only be performed by MDs. Neurofeedback would then immediately become essentially unavailable anywhere. Is there an entitlement here by the public, one similar perhaps to the entitlement to immediate care in an emergency? Is there an intrinsic right to access to medical care, for example? We do recognize a similar right when it comes to the legal realm. If a person is accused of a crime and cannot afford a lawyer, he will be given the support of a publicly paid attorney. Does a medical need not rank higher than a legal one?

It can be argued that in the evolution of human rights, the right to access to medical care might well come to be recognized. And if such a right exists, then it cannot be conditioned on the size of the person’s bank account. It would then follow that any policy that arbitrarily restricts care should not pass muster. An example would be what happened with Helen Irlen, when optometrists tried to put her out of business for practicing “their” trade. This campaign was not launched because optometrists thought that they would deliver the service more competently. They were trying to make the service unavailable, and that should not be ok.

Even the issue of efficacy is not paramount in this instance. Efficacy is first and foremost an issue for the licensed professions. License confers a legal privilege, and that carries certain obligations. The shaman and the healer, just to make the most obvious case, do not have the burden to prove efficacy. Their rituals could be nothing more than a vehicle for mobilizing the placebo effect, and that would be sufficient. Similarly, efficacy does not have to be demonstrated for nutritional supplements. The rights here are those of the end user in either case. On the one hand, the client needs to be protected from licensed people abusing their station with bizarre claims, but on the other hand the client should have full latitude to take advantage of new findings in the healing arts.

In Helen Irlen’s case, the issue of efficacy resolves itself trivially. No client would undergo a significant expense until efficacy with that person specifically had been demonstrated (as can readily be done with overlays). In neurofeedback as well, progress is observable early on in training. In both cases, the client is maximally empowered.

Finally, then, the Good Samaritan Model may be the basis for arguing that the right of the public to access to neurofeedback should not be arbitrarily restricted.

Leave a Reply