For a century, neurology has measured decline after it arrives. We are building the instrument that measures it before.

Modern medicine was built for a world in which measurement was expensive and clinicians were scarce. Patients were seen once a quarter, graded on rating scales read aloud in an examination room, and treated only once a diagnosis could be pronounced. This arrangement produced the 20th‑century hospital and most of the drugs we still prescribe. It also produced a habit of mind that we have come to accept as the nature of medicine itself: that care is a response, and that the nervous system in particular can only be known in retrospect, from the shape of its collapse.

Nowhere is the cost of that habit more visible than in neurological drug development. Pharmaceutical trials in Parkinson's disease are still adjudicated on the MDS‑UPDRS, a clinician‑scored rating scale with rater‑to‑rater variance measured in whole points and a detection floor well above clinically meaningful change. Sponsors spend hundreds of millions of dollars over five‑year timelines to power studies against an endpoint that is subjective, episodic, and blind between clinic visits. When a drug fails, it is often impossible to say whether the molecule was inert or the instrument simply could not see. This reality is an endless trap, prolonging the time it takes to understand if a drug is viable. The endpoint, not the disease, is setting the pace.

We believe this is the problem to solve first, and that the instrument to solve it with is already in every household. A consumer webcam (ie. phone, desktop) captures enough signal to quantify the involuntary motor behaviour of the nervous system, if the signal is handled correctly. Aethel's resampling engine takes arbitrary video at arbitrary frame rates and returns a uniform 25 Hz basis that is invariant to device, codec, and sensor. On top of that basis we train a neural ODE-based kinematic physics engine whose only purpose is to read the nervous system against its own prior state.

"We cannot know a disease until we have seen it in all its phases."

— Thomas Sydenham, Observationes Medicæ · 1676

Over the past year we have enrolled and processed 744 healthy participants through the full Aethel pipeline, producing the normative corpus against which every subsequent cohort is read. Enrolment for our first clinical study — a 30‑patient, 90‑day Parkinson's pilot with a partnering movement‑disorders centre — is now open, with a pre‑registered concordance target of r ≥ 0.85 against the MDS-UPDRS Gold Standard. Neither the corpus nor the clinical access is available off the shelf.

What we are building is not an application, and not a diagnostic. It is a measurement standard. Ideally, a unitary record of nervous‑system state that begins as the endpoint layer for pharmaceutical trials and scales, over the decade, into the annual nervous‑system examination for the 50M+ adults in the United States who will enter their seventh decade during it. The shift from reactive to predictive medicine will happen in roughly the time it takes a generation of clinicians to retire. Aethel is what that shift will be measured on.

We are a small team. We are deliberately institutional in posture. We correspond.