Platform
One engine that turns messy cancer signal into decisions.
HMT's software reads the reports every cancer patient already generates, places the findings on recognized clinical scales, and helps get the patient in front of the right doctor. Every time it runs, it adds to a record of what happened next.
Three layers, built in order
Detect, navigate, learn.
Detection earns its place first, navigation creates the network, and the registry is what makes the whole more valuable over time.
Detection
Reads what every cancer patient already generates, starting with radiology reports, and turns it into structured clinical information with the source sentence attached.
Detection layerNavigation
Takes those findings and suggests who should see the patient and where, across health systems. The clinician decides; the patient chooses.
Navigation layerRegistry
A prospective, outcome-linked record of what was found, what was decided and what happened next, with identifiers removed. It helps the software improve over time.
Registry layer
Design principles
Built as clinical decision support.
It shows its work. Every suggestion comes with the exact sentence that triggered it, so a clinician can check the basis in seconds rather than take the software's word for it.
The clinician decides. The software suggests a next step and nearby qualified specialists. It does not make the decision, and it does not contact patients on its own.
Neutral by design. Suggestions are driven by clinical need, specialist capability, distance and coverage, and the criteria are shown alongside them.
Minimum necessary data. Patient information is kept only as long as it is needed to help refer the patient, and identifiers are cleared before anything enters the registry.
Starting where the need is sharpest. HIPEC and peritoneal disease are where we start: a large group of patients who could benefit from a specialized operation and are too often never referred. The same engine extends to other cancers.
Keep reading
How it works
From a new finding to a clinician-approved next step, and how a deployment starts as a silent pilot.
Detection
What the engine reads today, what it produces, and the imaging track.
Registry
The prospective, outcome-linked record that builds itself as the system runs.