Veterinary radiograph workbench
Radiograph review and record keeping for equine vets.
EquiRay gives a vet an organised study instead of a folder of files: every view classified, every gap in the set named, image quality scored, and the vet's own findings carried through to the people who need them.
In short
View.Measure.Record.Track over time.A workbench for qualified vets.
The boundary
EquiRay measures. It does not diagnose, and it will not tell you what a change means.
The strongest verdict the software will ever reach on its own is that something needs review. A veterinarian accepts, and a veterinarian decides everything after that.
That is the design, not a limitation being worked around. A finding nobody checked is worse than no finding, because somebody will act on it. So the software organises, checks, measures what you point it at, and carries what you wrote — and the veterinarian judges.
How it works
Four stages, all of which we will say out loud
There is nothing in here that needs a euphemism. This is the whole shape of the product.
01
Ingest
Files arrive and become a study for one horse. Personal information is stripped on the way in, each file is checked to be a radiograph at all, and provenance is recorded for every one of them.
02
Audit
The set is compared against the repository standard. Each image is given an anatomical view and an image quality score, and anything the standard expects but the folder does not contain is named.
03
Read
The veterinarian works through the study in a reading room built for it, measures what they need to measure, and records findings against the study, with their name on them.
04
Carry
The findings and the veterinarian's own written report feed the buyer's risk read, so what the vet said travels with the horse instead of sitting in an inbox.
Today
Everything on this list runs, with tests behind it
Not a roadmap. This is the workbench as it stands, which is why there is nothing on it about an opinion on an image.
Study assembly
A folder of films becomes an organised study for one horse, with provenance recorded for every file.
View classification
Which anatomical view each image is, so the set can be checked against a standard rather than eyeballed.
Image quality scoring
Whether an image is good enough to work with at all, scored before anyone spends time on it.
Measurement
Angles, ratios and lengths, computed by code from landmarks you can drag. Angles and ratios are scale free, so the foot angles work on any film. A length is given in millimetres only where the film carries a calibration, and is marked uncalibrated where it does not, rather than printing a pixel count that looks like a measurement.
Two studies, side by side
A horse filmed at a repository and filmed again months later. The arithmetic between the two, in the unit both numbers were already in. A measurement taken on one study and not the other is reported as not measured, naming the study that has it, because not measured is not zero.
A series, in the order it was taken
Three sittings or ten, laid out in order with the interval between each, the difference between each, and every hole in the chain named. What it will not do is tell you where a run of numbers is heading. That sentence is a clinical judgement and it is yours.
Repository audit
Which views the standard expects and which the folder does not contain. A missing view is a risk flag in its own right.
Personal information stripped on ingest
A film carrying a client's name does not carry it into our system. Removed on the way in, and pinned byte for byte against a second implementation.
Viewing and measurement
A reading room built for the work, with the measurement tools in it. The veterinarian reads, measures, and records findings against the study. The finding is recorded as theirs.
Horse and client records
Structured findings against a horse, kept so that the same horse read twice a year apart is one record rather than two folders.
The write-up reader
It takes the veterinarian's own written report, in prose, and pulls out what was said so the buyer's risk read can use it. It works on the words, not on the pictures.
Not on offer
Each of these is a decision, not an oversight
This is the part you can hold us to. If anything in the product ever contradicts a line here, that is a bug and we want to hear about it.
- No verdict on the horse
- The strongest thing the software reaches on its own is that something needs review. It does not go further, and there is no setting that makes it go further.
- No machine finding
- Every finding on this platform was made by a veterinarian and is recorded under their name. The software never originates one.
- No opinion about an image
- We have not built one, and that is on purpose rather than by accident. It stays unbuilt until it can be earned on work a veterinarian labelled, because a machine finding that nobody checked is worse than no finding at all.
- No clearance
- A complete set with nothing recorded against it is exactly that: a set with nothing recorded against it. It is not a statement about the horse, and EquiRay will never present it as one.
- No number for how often we are right
- You will not find a percentage anywhere on this site. A figure like that needs a study behind it, made at the time the claim is made, and we have not run one.
- Never a veterinary diagnosis
- EquiRay is decision support for a qualified veterinarian. The clinical judgement is the veterinarian's, the examination is the veterinarian's, and neither of them moves.
What we are building
Measurement is the foundation, not the destination. Here is what we are building on top of it.
- Findings on the film itself
- A model trained on vet-labelled studies that proposes findings on a radiograph, not just landmarks. It will still be a proposal: a finding nobody checked is worse than no finding, so a veterinarian will accept or reject every one.
- Change, explained in plain English
- Two studies of the same horse, months apart, with what changed and by how much — written so an owner can follow it, with the measurements underneath so a vet can audit it. Today the software shows you both and refuses to characterise the difference.
- Recovery, followed over time
- A series in the order it was taken, so a horse coming back from an injury can be followed rather than guessed at, and so a change that is going the wrong way is visible early.
- Beyond radiographs
- Ultrasound and endoscopy alongside the films. Neither is supported today — ultrasound is refused at upload on purpose, because accepting an image the software cannot read would be worse than declining it.
- Movement and structure together
- StrideVision measures how a horse moves; EquiRay organises what is inside it. Read together they describe a horse more completely than either does alone, and that pairing is where both products are heading.
None of this is in the product today. It is stated here as intention, not capability, and this section will shrink as the one above it grows.
Early access
We would rather build this with a handful of vets who will tell us plainly what is wrong with it
EquiRay is in development. If that is of interest, tell us what you image most and how the repository reaches you today.