Bringing neurofeedback into a mental health practice means acquiring a working model by means of which all the clinical phenomenology can be reframed in a psychophysiological perspective. Clinical decision-making then emerges largely out of that framework. As the training proceeds, clinical observations are interpreted in terms of that framework and lead to fine-tuning of the clinical strategy. There are two feedback loops here. One involves the client and the feedback signal. The other involves the client and the clinician. The importance of the latter has increased over time as the techniques have strengthened in their impact. Within-session changes in physiological state have to be attended to promptly to steer the training in a propitious direction. This responsiveness, which is observed fairly typically, means that the burden of clinical decision-making has shifted from being a rather freighted decision at the outset to a continual, iterative process that is itself feedback-guided toward the desired objectives. This places the principal burden of competent guidance upon the clinician—more so perhaps than with any other approach.

One proceeds from the basic orientation that the brain must satisfy all of the criteria of a feedback control system. We think in terms of hierarchies here. Firstly, the CNS must assure its own unconditional stability. Allied with this concern is the ability to contain behavioral disinhibition. Second, it must manage set-points of activation of different functional domains. Thirdly, it must arrange for the smooth integration of these functional domains to meet the challenges of life. Clinical targeting then follows this same hierarchy. Promoting brain stability is the first objective. Training for better management of states of activation is the second. Functional integration to manage localized deficits is the third.

Our regulatory networks are also seen as hierarchically organized, with the brainstem as the head of the hierarchy in terms of the organization of timing in the orchestration of cerebral processes. Hence the first objective in terms of state regulation is the domain of arousal regulation and of the sleep/wake cycle. We no longer try to characterize people in terms of high and low-arousal tendencies. That is no longer relevant to clinical decision-making. But understanding how people function in the arousal domain yields predictors about how they will react to the training. If these expectations are not met, then one either adjusts one’s approach or one’s understanding of the client.

Arousal regulation is primary in our considerations for a number of reasons, but a key factor is that it is a good observable for us. We can see readily how people react in the arousal domain. This is key because if we are able to move the person to a well-regulated state in the moment, then we can be sure that this will also yield the desired outcome over the longer term. The two are highly correlated. It even holds true that when we optimize reinforcement parameters for arousal regulation we are at the same time optimizing conditions for other regulatory functions that are not so readily observable. Further, if we have any other cues regarding the person’s self-regulatory status, these also serve as indices for guiding the training appropriately. We take advantage of a kind of unitary quality to self-regulatory competence with respect to core regulatory functions.

The second priority is affect regulation, because of the intimate relationship between the limbic system and the brainstem. This also ties us in preferentially to right hemisphere function, so already at this point laterality issues are paramount. Failures associated with the right hemisphere are potentially the most catastrophic for people, and hence demand our early attentions. The right hemisphere is also the earliest to develop, so our training hierarchy aligns with that of early childhood developmental stages.

Addressing left hemisphere function is only the third priority, and that also brings in the left pre-frontal region. This ordering of priorities is somewhat ironic, given that most of us got started in this field with a focus on ADHD. Further, responsibility for a lot of self-regulatory deficits has been assigned to the frontal lobe. But matters are not as they seem. It has been clear to us for a good many years that emotional disregulation lies at the core of much that is labeled ADHD. The high overlap of ADHD with oppositionality and Conduct Disorder attests to this. It is the emotions that govern our attentions, and hence deserve priority in our hierarchy.

As the above is beginning to make clear, we organize our thinking in terms of principal regulatory axes: the top-down axis (with the brainstem at the top of the hierarchy!), the left-right division, and the front-back axis. Each of these constrains the framework in specific ways. Within these divisions, we place the priority on training bottom-up control versus top-down (and in this real-world picture the brainstem is at the bottom, where it belongs!), and we attend to the trophotropic division before we attend to the ergotropic. This is in line with the new findings regarding our resting state networks, namely that the organization of our resting states is a good predictor of our functional competences.

This is also in line with our biofeedback heritage, in which it was recognized early on that the quality of autonomic nervous system regulation was the key to good function more generally. In the early days of neurofeedback it was certainly the view that EEG training would simply complement peripheral biofeedback where the latter did not offer good access—executive function, specific cognitive function, working memory, visual and auditory processing, etc. Now it turns out that EEG feedback may be an elegant and highly efficient means of achieving autonomic nervous system regulation as well, in which case it needs to be placed early in the hierarchy of clinical concerns.

Historically, our migration from the standard SMR-beta training that we taught for many years to what is now predominantly training in the infra-low region of EEG frequencies took over ten years, over which time our clinical competences were enhanced particularly with regard to emotional regulation, autonomic regulation, and interoception. Along the way, clinical priorities needed to be re-ordered. Our approach has by now diverged sufficiently from prior practice, as well as from common practice within the field, that it is called the Othmer Method. The distinctive features are that it utilizes bipolar montages in the principal constituents of the training program, and that reward frequencies are individualized. A total of five bipolar montages constitute the complete set from which a starting protocol is constructed.

The set of five protocols allow us to challenge the regulatory role of each quadrant of cortical real estate, with the fifth protocol devoted to inter-hemispheric training to promote cerebral stability. Our anchor sites are T3 and T4. These sites monitor the multi-modal association area of the anterior temporal lobe, and interact with the limbic system and nearby insula. The T3-T4 placement is used standardly in application to instabilities across the board, including vertigo, panic, asthma, migraine, seizures, and bipolar excursions. The other standard placements are T4-P4, used for physical calming; T4-Fp2, used for calming emotional reactivity and rage; T3-Fp1, used for mental calming and improved executive control; and finally T3-P3 for awareness of detail and related sensory processing issues. The latter category rarely rises to the level of a priority, so as a practical matter five starting protocols reduce to four.

In practice, the clinical hierarchy is implemented as follows. The primary target, as already stated above, is cerebral stability. Given the high prevalence of instabilities in our clinical population, T3-T4 is the dominant starting protocol. This general rule may be trumped in the case of developmental delay, early trauma, or other severe disorder traceable to early childhood development. In these cases the high arousal to which these nervous systems are driven becomes the paramount consideration. It can itself be a contributor to cerebral instability, so even in the event that instability is the primary concern (as in autism with seizure risk), the best approach may nevertheless be to calm the right hemisphere as a first priority. The starting protocol in these cases is T4-P4. If neither instabilities nor developmental issues are prominent, then matters are treated as more ordinary disorders of disregulation with a starting protocol that combines right parietal with left frontal training. In such cases T3-T4 remains the starting protocol as the best means for finding the optimum training parameters. The details are given in The Protocol Guide (Ref)

In terms of reinforcement frequency we have found our way over the years to the infra-low regions of EEG frequencies (i.e., < 0.1 Hz). Here we are simply tracking the slow cortical potential (SCP) in its temporal migration. This cannot be represented well with a narrow spectral filter. But at such low frequencies one cannot have both a narrow-band filter and real-time information. So we are actually seeing this signal within a broad signal bandwidth, and it does not appear at all sinusoidal. On the other hand, we have found clients to be highly sensitive to the particulars of the passband, and in this respect the training does resemble traditional frequency-based training. The adjustment of the band edge is the means by which the clinician optimizes the response for each individual during each session.

Unfortunately this technical approach violates all of the rules we have come to associate with traditional neurofeedback. First of all there is no event here that could be rewarded in the traditional mode. There is not even a goal that could be defined, so there is no threshold. There is no better and there is no worse. We are looking at a differential signal with broadly cyclical properties, after all, and a rising signal cannot be considered more virtuous than a falling one. (If we had exchanged electrode placements we would have the reverse polarity.)

The process is therefore one in which the brain is challenged toward a better self-regulatory status without our being able to steer or direct the process in any way at the level of the EEG. We simply provide the brain with information on its own slow cortical potential, and it reacts to that information. No instructions need to be given. We have to conclude that the challenge is built into the process. The brain recognizes that it is the author of the signal on the screen, and it naturally assumes responsibility for that signal. An error function inevitably prevails between the signal on the screen and the interpretation the brain gives to that signal in terms of its own ongoing activity. The minimization of that error function ongoingly constitutes the relevant challenge to the brain.

Now this view of physiological feedback may be strange to us in the neurofeedback community, but in fact peripheral biofeedback has been living with this reality since the beginning. We would describe Heart Rate Variability training in the same way, with the additional factor that in the latter case the instruction may be given, for what it is worth, to increase the excursions in heart rate. A similar bias is implicit in many of our feedback games, in that larger signals are still favored as a carryover from the earlier days of SMR-beta training. Subject to that bias, our clients would be motivated to enhance the signal amplitude, which translates into greater excursions of the underlying variable, the SCP. So it could be said that we are promoting increased dynamics in the SCP. But there are other feedback modes which harbor no such bias, and we have clients who cannot follow instructions. We must conclude that the challenge is imbedded in the process itself, and is not imposed by the particulars of our mechanization.

We have developed a training method that combines some of the characteristics of traditional frequency-based training with features of traditional SCP-type training and of conventional biofeedback. The combination appears to be an improvement on each of its antecedents, for purposes of state regulation, with respect to breadth of clinical footprint, accessibility by various target populations, scope for skillful guidance by the clinician, rapidity of clinical response, and engagement of the client in the process.

References
1. Othmer, Susan F., (2008). The Protocol Guide, www.eeginfo.com

Siegfried Othmer, PhD
drothmer.com

This entry was posted on Friday, April 8th, 2011 at 7:21 pm and is filed under Neurofeedback. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.