How a wealth of findings can become an understandable explanation – and what our four-axis model aims to contribute

2 November 2026 · @Andreas Müller

The report lies on the table: restlessness, outbursts of anger over little things, great difficulty with change, conspicuous scores on the attention test, plus measures from the brain function analysis. Then the mother asks: “Is it all connected?”

We hear this question in almost every assessment. A report can be thorough and still leave it unanswered. It places symptoms next to test scores and test scores next to EEG findings. The family receives many precise individual pieces, but not yet an explanation.

This experience gave rise to the four-axis model at the BrainARC-Ostschweiz practice. It is meant to help trace many findings back to a few underlying processes. It has not yet been proven. This article explains what it sets out to do and where its limits lie.

The diagnosis stays, but it comes later

A diagnosis provides classification. It is often the prerequisite for insurance cost approvals, school support measures and medication. The model does not change that.

But it says little about why these particular difficulties occur together in a given person. Two children with the same diagnosis can function very differently. And a child with three diagnoses does not necessarily have three separate problems.

That is why, in the report, we first ask which processes explain the findings. We then check which diagnosis fits and whether its criteria are met.

Four domains and two factors that affect them all

The model combines two research approaches. The Research Domain Criteria organise the observations (Insel et al. 2010). Predictive processing provides the explanation: the brain continuously forms expectations and compares them with new information (Clark 2013). This yields four functional domains, which we call axes.

Axis	What it is about
Stimulus processing	How much gets through, and how much weight is it given?
Self-regulation	Do plans and routines hold up, and can they be changed when the situation changes?
Emotional reactivity	How does a person interpret and regulate bodily signals and feelings?
Social perception	What assumptions does a person form about what is going on in others and in themselves, and can these be corrected?

For social perception, we draw on mentalising research (Luyten et al. 2020). Two further factors act on all four domains.

Wakefulness and activation. The technical term is arousal. We read it first because it can shift everything at once. One example: in everyday life a child may appear mainly restless and, during a long, low-stimulus test, increasingly miss tasks. According to the vigilance model of Hegerl and Hensch (2014), movement keeps an under-aroused brain awake. If movement is removed, wakefulness drops. This is an interpretation that must be tested in each individual case.

The environment. Self-regulation and emotional regulation develop and are sustained in relationships. Images have an effect too: someone seen in the family or at school as “the difficult child” is often still treated that way long after something has changed. That is why, in the model, work with parents and schools does not count as an add-on, but as an equally important point of intervention.

Three questions for each domain

Imagine each axis as a camera. We ask each one the same three questions.

1. How sharp is the image? This means how certain the brain currently is in this domain. Sharp does not mean correct: someone can be very sure they know what others are thinking and still be wrong.
2. How quickly does the autofocus readjust? This means how strongly new experience changes the existing image. Too slow is as much of a problem as too fast.
3. How much power does the camera use? This means the effort someone puts into this performance. An image can be sharp and the battery still run flat after a short time. The distinction between control exerted and its costs draws on Shenhav et al. (2013).

The first two questions yield four typical patterns:

	Slow adaptation	Rapid adaptation
Sharp image	Rigid persistence: routines or assumptions remain in place even when the situation has changed.	Overreaction: every new impression changes the state strongly, for example with high sensitivity to stimuli.
Blurred image	Driven by the strongest stimulus: plans carry little weight, and learning from experience is slow.	Constant switching: every new piece of information is taken on, but nothing builds up.

High effort cuts across these patterns. It can occur alone or accompany and mask another pattern.

These patterns are not types of people. A child can react rigidly at school, be flexible at home and flip back and forth under stress. That is why the report always also describes in which situations a pattern occurs and when it eases.

An example: a good test, a lost evening

Lena is fictional, but the picture is typical. The 15-year-old has good grades and is exhausted after every school day. At home, the smallest thing is then enough to start an argument. ADHD is suspected, but the attention test initially goes well. Only towards the end of the long task does her performance decline.

Taking only the overall score, the matter would be settled quickly: test unremarkable, so no problem. With the three questions, the picture looks different. The sharpness is fine; the power consumption is too high. This fits with her own account, her parents’ observations and the drop in performance under sustained load. Individual measures from the brain function analysis may fit this too, but that proves nothing.

In the consultation we then say something like: “Our current understanding is that Lena is using a great deal of strength to compensate for what costs others less.” The first consequence is relief: more recovery, better sleep, simpler strategies. Even more concentration training would probably be the wrong direction here.

Whether ADHD is also present is examined separately according to the criteria. If it is confirmed, medication in line with guidelines can be considered. Without the diagnosis, exhaustion is no reason for a stimulant.

What this is meant to change

The language. Technical terms such as precision or update rate do not belong in the report. Instead of “Self-regulation is rigid”, we write something like: “In unpredictable situations, the child finds it hard to change familiar routines. When changes are announced and the child feels safe, this works much better.” This keeps it visible that things can change. The perspective of those affected and their strengths also belong in the report.

The sequence. If an under-aroused child also clings rigidly to routines, the recommendations contradict each other: one calls for structure, the other for change. We propose first stabilising wakefulness and everyday life, then gradually practising change. This can be checked: if the clinging decreases with stabilisation, it was probably compensation. If it remains, it needs treatment of its own.

Something similar applies to adolescents who see hostile intentions when tense and are empathetic in calm moments: the tension must come down first. Then work on alternative explanations can take effect.

Not the medication. Its indication continues to follow the diagnosis and medical judgement. The model justifies neither medication without a diagnosis nor the choice of a particular substance.

Rules against the all-too-neat explanation

A good story convinces even when it is wrong. That is why fixed rules apply.

• A domain counts as affected only if findings from at least two different sources support it. One of these must be clinical: interview, questionnaire or test.
• The brain function analysis alone is never sufficient. A conspicuous value demonstrates a deviation from the age norm, but not yet its meaning.
• What cannot be classified remains open and is stated as such in the report.
• ADOS-2 and ADI-R remain the gold standard for an autism diagnosis. The model does not replace them.

Practice has already corrected the model. We applied it in 83 reports between May and July 2026. Social perception did not appear once as an affected domain. Almost all reports concerned children and adolescents, and no validated instrument is available for younger children. Even so, our rule was also too narrow, and we changed it. Whether the new rule is better must be shown in other cases.

What this text does not show

The four-axis model is a proposal from clinical practice and not a validated procedure. The 83 reports show that it can be applied and where it has gaps. They come from a single centre. An independent second assessment is still pending.

Above all, the most important question remains open: does the model lead to better decisions than a careful conventional assessment? For the first treatment step, that assessment would often arrive at the same approach. The neurophysiological links and the recommended sequences are testable hypotheses. The model is also not suitable for self-diagnosis.

The accompanying scientific manuscript is in preparation and has not yet been peer-reviewed.

More on 10 December

Two of the questions lead directly to the Biomarker Workshop 2026: How much gets through, and what does it cost? Sensitivity and stress as connecting dimensions across ADHD, autism and PTSD are the focus on 10 December 2026 at the Paulus Akademie in Zurich. The workshop is aimed at professionals; participation is possible on site or online.

Register now

Sources

• Clark, A. (2013): Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences 36(3), 181–204. DOI 10.1017/S0140525X12000477
• Hegerl, U.; Hensch, T. (2014): The vigilance regulation model of affective disorders and ADHD. Neuroscience & Biobehavioral Reviews 44, 45–57. DOI 10.1016/j.neubiorev.2012.10.008
• Insel, T. et al. (2010): Research Domain Criteria (RDoC): toward a new classification framework for research on mental disorders. American Journal of Psychiatry 167(7), 748–751. DOI 10.1176/appi.ajp.2010.09091379
• Luyten, P.; Campbell, C.; Allison, E.; Fonagy, P. (2020): The mentalizing approach to psychopathology: state of the art and future directions. Annual Review of Clinical Psychology 16, 297–325. DOI 10.1146/annurev-clinpsy-071919-015355
• Shenhav, A.; Botvinick, M. M.; Cohen, J. D. (2013): The expected value of control: an integrative theory of anterior cingulate cortex function. Neuron 79(2), 217–240. DOI 10.1016/j.neuron.2013.07.007