Radiology research,
built on provenance.
Melkorka Labs is a radiology research company. Our platform, Melkorka, is built for quantitative radiology: a native viewer, a deterministic study orchestrator, and a computation kernel — designed together so that every image, every model, and every number can be traced to the exact computation that produced it.
The Viewer
A native macOS DICOM viewer rendered on Metal. Multi-planar reformatting, freely oriented oblique slicing, linked viewports, and computed models drawn directly over the source images — interactive, at full fidelity.
Melkorka Loom
The orchestration layer. A study is expressed as a deterministic program over a frozen cohort; Melkorka Loom runs it, records how every thread of the fabric was woven, and keeps the full provenance of each answer.
The Kernel
The computation engine. It builds tissue models and derived artifacts from imaging data, addresses every result by the content of the request that produced it, and never computes the same answer twice.
Method
Results are addresses
Every artifact in Melkorka is content-addressed: the identity of a result is the computation that produced it. Two studies that ask the same question of the same data get the same answer — byte for byte.
Numbers from structure
Our research interest is measurement anchored in the tissue itself — numbers defined by physical transitions in the image, rather than by thresholds chosen after the fact.
Evidence over plausibility
The platform is developed against invariants: fuzz-tested state, telemetry on every commit, and oracles that prove a behavior rather than demonstrate it.
Melkorka is the research platform of Melkorka Labs, in active development. It is not a medical device and is not cleared for clinical use. If the work interests you — as a researcher, radiologist, or collaborator — we would like to hear from you.
halldoro@gmail.com