Why we started with the pelvis

Picture the moment this product exists for. A pelvic-floor physiotherapist has two, maybe three minutes to explain a finding to a patient. The anatomy in question is internal, intimate, and impossible to point at. The patient has never seen it, cannot picture it, and their understanding of it — the thing that decides whether they do the exercises — is about to be built entirely out of words.

Most anatomy software was not made for that moment. It was made for a different one: a student at a desk, preparing for an exam, with the whole body to cover and unlimited time. Those are honourable tools, but the job shapes the tool. Built for coverage, a general atlas treats every region with the same democratic shallowness — and the pelvic floor, which is where our users live, tends to arrive as a single fused shape with one label on it.

Depth where the work is

We inverted the priority. One region, modelled to the depth the specialty actually uses, for the conversation itself.

Concretely: the levator ani in our model is not one muscle. Puborectalis, pubococcygeus and iliococcygeus are separate, selectable structures, alongside the obturator internus, the pelvic organs, the support ligaments and the pudendal nerve. Any of them can be isolated, faded, rotated to the angle that makes the explanation obvious — because "which part, exactly?" is the question a specialist answers all day, and a model that cannot distinguish the parts cannot help answer it.

Open this structure in the viewer

Built for the appointment, not the lecture hall

The second inversion is about time. A tool used mid-appointment has a budget of seconds: open the right scene in one click, switch between professional terminology and plain patient language without losing your place, turn the screen toward the patient and talk. Clinical presets exist for exactly this — the scene you need, ready before the patient has finished sitting down.

Nothing to install matters for the same reason. It runs in the browser on whatever laptop or tablet is already in the room. No IT ticket stands between a clinician and a better explanation.

Why pelvic health specifically

Because this is where the gap is widest. The anatomy is invisible and genuinely difficult to describe. Patients arrive with symptoms they have sometimes been embarrassed about for years, and the single thing that most changes their engagement is understanding what is happening inside their own body. Clinicians deserve a better instrument for that conversation than a plastic model and a pen — and the specialty is underserved precisely because it is specialised, which is the opposite of how a coverage-first product allocates its effort.

The technology underneath is a general anatomy engine; nothing about it is pelvis-only. Pelvic health is where it starts, not where it ends. But starting narrow is the point: depth is the product, and depth is only possible one region at a time.

What taking a region seriously means underneath

Depth is not only polygon counts. The unglamorous foundation is an ontology: a fixed list of the structures the region contains, each with a stable identity, a professional name, a plain-language name a patient can hold onto, and a position in the hierarchy — this muscle belongs to this group, attaches here, sits beside that. Ours currently holds thirty-five structures for the female pelvis, every one selectable, every one with terminology in two registers.

Stable identity sounds like a database detail until you see what it buys. It means a saved scene reopens exactly as it was left, months and versions later. It means a clinical preset can name its target and never miss. It means when the commissioned model replaces our placeholder geometry, everything built on top keeps working, because the structures' identities outlive any particular mesh. And it means the patient-language layer is not a translation bolted on afterwards but a first-class property of every structure — switching registers mid-explanation is instant because both names were always there.

None of this is visible in a screenshot. All of it is visible in the third minute of an appointment, which is where this product is meant to earn its keep.

If you are reading this as a potential collaborator

This post doubles as our introduction to the medical artists, academic partners and early clinicians we write to. The short version of how we work: structure identities are stable and versioned, commissioned geometry is licensed work-for-hire with the paperwork done properly, and every piece of clinical content ships with a visible review status — we have written separately about that. If you make anatomical models, review clinical content, or run a pelvic health programme and any of this sounds like your kind of rigour, we would like to hear from you.

And if you are a pelvic-floor physiotherapist wondering whether the model above is real: it is, and the demo is open. Early access is opening in stages — the form on the home page is the way in.