Somewhere right now, a small team of very capable people is rebuilding infrastructure that already exists.
Not because they want to. Their ambition is building a biomarker, or a closed-loop stimulation protocol, or a decision-support tool that might finally tell a clinician which treatment to try first. But to do so, they need clean signal from a head that moves, electrodes a nurse can apply in ninety seconds, real-time software that doesn't drop samples, and eventually a technical file that survives a notified body audit. None of that is their idea. All of it stands between them and their idea. It will take eighteen months and a funding round they will never get back, and at the end of it they will own an acquisition stack that is merely adequate.
I have watched this happen more times than I can comfortably count. It is the most avoidable waste of talent in our field and a reason for good ideas to die, and it is why we opened up our technology partnership program at ANT Neuro.
The need is not short of ideas
Brain and mental health is where medicine is most visibly struggling. The WHO now estimates that around 332 million people live with depression, and that even in high-income countries only about a third of them receive treatment for it (1). Waiting lists are long and getting longer. And in psychiatry, unlike most of medicine, treatment is still largely by trial and error: a diagnosis reached through conversation, a first-line treatment chosen by convention, and months of waiting to find out whether it worked. For a substantial share of patients it doesn't — and then it starts again.
Somewhere right now, a small team of very capable people is rebuilding infrastructure that already exists.
It would be wrong to say the technology hasn't moved. (Navigated) TMS is delivered to patients every day. aEEG sits at the cotside in neonatal units that could not realistically have run it ten years ago. Research-grade EEG has left the shielded room and combines readily with other modalities. We have had a hand in all three.
What is arriving too slowly is the layer above the instrument. Which cortical target for this particular patient, rather than one defined by a scalp measurement. Which electrophysiological signature predicts that this person responds to that treatment. Which stimulation parameters, and how to know at week two rather than week twelve whether they are working. How a unit without a neurophysiologist on site acts on a trace at three in the morning. Those are the questions that decide whether a patient actually gets better — and most of the answers still live in papers, in single-site pilot cohorts, and in software that runs on one laptop in one lab.
So the gap is not an absence of instruments. It is the distance between a promising method and a method that is validated, cleared, supported and genuinely used. And that layer is largely not ours to build.
What we do ourselves — and where it stops
Our neonatal monitoring system exists because aEEG was too cumbersome to set up in many of the NICUs that most needed it, and making it routine changes how fast a seizure is caught in a cooled newborn. Our neuronavigation system exists because stimulating a brain without knowing where the pulse lands produces exactly the variability you would expect, and guiding it from the patient's own anatomy or functional map makes TMS reproducible between sessions and between operators. Those are real diagnostic and therapeutic contributions, delivered in real clinics.
But the set of open problems in brain and mental health that we can address end to end ourselves is a small fraction of the challenge at large, however well we execute. The number of indications, populations, methods and care settings that need attention far exceeds any single company's engineering capacity, clinical depth, or attention span. Even the ceiling on our own products is set largely by that method layer — and we are not the ones who will produce most of it.
That is the honest arithmetic. We envision that most of what we can contribute to the rest of the field, we will contribute through other people's products.
What we can put on the table
What a partner builds on is not a box with an export button. It is:
- Medical- and research-grade EEG systems, electrode caps, and neurodiagnostic and neuronavigation platforms already in clinical and research use
- Integration at the level a serious application actually needs — data, timing, control, and software interfaces
- Scientific and clinical expertise, including people who have taken devices through MDR and FDA clearance and know precisely where it hurts
- Global sales, service, and distribution — a route to clinics and labs that would otherwise take years to build
- A relationship built to last, because a clinical product that depends on a supplier needs that supplier to still be there in 2035
What partners bring is what we cannot: deep knowledge of a specific indication, clinical know-how earned by sitting with patients, the specialised algorithms and workflows that turn a signal into a decision, and the conviction to keep going when the first study is ambiguous.
We can only solve a small number of these problems ourselves. We want to be part of solving a great many more.
The wider point
If you are building a digital biomarker, a brain-computer interface, a neuromodulation product, or a clinical biosignal application, we should rather talk early than late. Early means before the architecture is fixed, before a hardware team is hired, before the acquisition layer is written into the roadmap and the investor deck. Afterwards, those decisions are expensive to unwind.
We won't be the right partner for everyone, and some conversations will end after the first call; perhaps with a referral to one of our own partners. If that's where we land, you'll have lost an hour and gained a reasonably well-informed second opinion and perhaps an introduction to one of our partners that may be more suitable.
But the ambition is larger than any single collaboration. The same foundation serves the neonatal unit, the psychiatric clinic, the rehabilitation ward, the research lab and the BCI startup — and every application built on it makes that foundation better for the next one. What I would like, ten years from now, is not a bigger catalogue with our name on it. It is a field where a great many of the tools that finally moved brain and mental health forward turn out to have been built on shared infrastructure, by people who spent their scarce years on the part only they could do.
We can only solve a small number of these problems ourselves. We want to be part of solving a great many more.
Details are here: ant-neuro.com/tech-partnerships. There's a form, but a message to me works just as well.
1) World Health Organization, Depressive disorder (depression) fact sheet: who.int/news-room/fact-sheets/detail/depression
Originally posted on LinkedIn.
