A scan that appears to identify the right treatment can be a powerful proposition for someone who has already tried several options. The important question is what using that information changes. A substantial brain-imaging study identified research groups and treatment associations, but did not test whether a scan-guided care pathway outperformed an appropriate alternative.
Key findings
801 participants: the clinical imaging sample with depression or related anxiety disorders, alongside 137 healthy reference participants.
41 circuit measures: the information used to derive six research biotypes.
250 participants: the treatment dataset used to examine associations with medication or behavioral-treatment outcomes.
15 participants: the smallest biotype group. These counts are research-group sizes, not population prevalence. [1]
Six research groups, not six new official diagnoses
Counts and simplified descriptions from Tozzi and colleagues, June 17, 2024. No prominent difference on the selected measures does not mean no illness. [1]
The study organizes a varied clinical sample into patterns worth investigating. That contribution does not depend on treating the groups as a replacement diagnostic system.
At baseline, 95% of the clinical sample was not receiving antidepressants, and none had a substance-dependent disorder. Those eligibility characteristics matter when considering complex private-care referrals.
Classification, prediction, and better decisions are separate tests
| Question | What an evaluation should establish |
|---|---|
| Does the measurement work? | Reliability, reproducibility, and interpretation |
| Does it predict something useful? | Performance in new patients and added value over existing information |
| Does using it improve care? | Better outcomes from test-guided decisions than a credible alternative |
A tool can reliably assign a category without improving treatment choice. It might predict poorer outcomes regardless of treatment, which is prognosis rather than evidence about which option works better.
The 2024 study examined treatment-outcome associations. It did not randomize patients to a scan-guided decision pathway versus standard decision-making. That leaves clinical promise and a further evaluation question side by side.
The treatment subset was not the whole sample
The 801-person clinical sample and the 250-person treatment dataset answer different questions. The study should not be described as successfully choosing treatment by scanning all 801 people.
Small biotype-treatment combinations also deserve proportionate uncertainty. A group of fifteen is not meaningless, but a few different observations could materially change a percentage within it.
The chart therefore uses counts rather than the supposed distribution of depression worldwide. Recruitment, eligibility, and the imaging framework shape the observed groups.
An unremarkable pattern does not invalidate symptoms
A scan captures selected features under specified conditions. Failing to find a prominent difference in those features cannot establish that distress is unreal.
Biological explanations can help people make sense of suffering. They should not become a test of whether someone deserves care.
The useful question is what information the result adds to symptoms, history, functioning, and the patient’s priorities. The measurement should complement an appropriate assessment rather than override it.
Personalization still needs a clinical comparison
A separate stimulation trial published in September 2026 tested specified imaging- and frequency-personalized approaches against sham. It did not establish an advantage on its primary depression outcome. This was not a replication or disproof of the six-biotype study. [2]
Its relevance is the broader principle: using individual information should be evaluated as a strategy, not assumed to confer benefit.
Imaging mechanisms, selecting a stimulation target, and choosing between medications are different applications. Evidence for one cannot automatically validate another.
More tests do not necessarily mean more useful information
Affluent patients may be able to arrange extensive investigations. That can help when a test addresses a relevant question. It can also make an impressive output available before anyone explains its role in the decision.
Ask what each possible result would change. If every result leads to the same recommendation, the added value needs explanation. If a result could lead to a major treatment change, its uncertainty deserves careful discussion.
This is not a cost-effectiveness calculation. It is a decision standard: expenditure should connect to a clinical question, interpretable information, and an appropriate action.
Complex patients need a relevant validation population
A result obtained largely before antidepressant treatment and without substance dependence may not transfer automatically to someone taking several medications or seeking integrated addiction and mental-health care. [1]
That does not establish that the approach could never help. It identifies a population in which the intended use needs evaluation.
For patients moving between countries, methods and interpretation also matter. Two providers using the general label brain imaging may not use the same protocol or produce equally validated decisions.
What to ask before purchasing a biomarker-guided plan
| Ask | Why it matters |
|---|---|
| Which decision will change? | Diagnosis, prognosis, and treatment selection need different evidence |
| Was this exact method studied? | A similar label does not establish equivalence |
| How similar were the participants? | Current treatment and co-occurring needs affect applicability |
| What is uncertain? | Categories should not conceal prediction limits |
| What follows each result? | The test needs a clear clinical purpose |
| How will progress be reviewed? | A prediction is not a guaranteed outcome |
These are proposed discussion questions, not a validated diagnostic instrument or a claim that THE BALANCE offers this test.
Patients should be able to understand why the test is recommended without needing to interpret technical brain maps themselves.
The next trial should test decisions, not only categories
A prospective test-guided trial could compare outcomes after biomarker-informed allocation with a credible alternative. It should examine benefits, harms, cost, and practical burden.
Independent evaluation at new centers would show whether the method travels beyond the original setting. Reproducing a category is useful, but the clinical aim is a better decision or result.
Longer follow-up also matters. A predictor of early symptom change may not predict sustained functioning or recurrence. Claims about durable recovery need observations on that timescale.
Information should remain understandable across the pathway
A patient seeing several professionals should not become the sole interpreter of a test result. Reports should distinguish established clinical information from research classifications and indicate how they informed the recommendation.
Where another clinician takes over, the reasoning and remaining uncertainty should travel with the report. A technically detailed document without an actionable explanation may add complexity instead of resolving it.
The proposed standard concerns communication and continuity, not an assumption that international patients experience a measured rate of testing errors.
What this evidence cannot tell us
The study does not establish six new official diagnoses, a commercial test’s cost-effectiveness, or improved outcomes from scan-guided care for all patients. It also does not invalidate other established clinical uses of imaging.
No source here provides a separate HNW prediction or treatment effect. Resources and professional status are not substitutes for the clinical characteristics used in validation.
The bottom line
Precision is a claim to test. The biotype research is a substantial contribution to understanding differences between patients, and its clinical translation remains a separate question. Ask whether purchasing the specific test will improve the decision being made, not simply whether its output looks scientific.
For journalists
Key distinction: six research biotypes were derived from 801 clinical participants; treatment associations involved a subset of 250.
Important caveat: this was not a trial showing that scan-guided allocation improved care over standard decisions.
Suggested attribution: THE BALANCE analysis of published brain-biotype and clinical-utility research.
Methodology and sources
This narrative appraisal separates classification, prediction, and clinical utility. The chart reproduces sample counts, not population prevalence. The 2026 stimulation study illustrates a different use of personalization and is not presented as a direct replication.
