The Pathology Network

In much of Kenya, a hospital that takes a biopsy can wait weeks for the result. There are too few pathologists, most of them in the big cities, and glass slides travel by courier. The Pathology Network (TPN), a Meru-based startup, set out to change that by digitising slides and having them reported remotely. I joined to build the software that makes that possible.

The problem

• Scarce specialists. Remote and county hospitals rarely have a resident pathologist.
• Slow logistics. Shipping physical slides and paper reports added weeks between biopsy and treatment.
• Isolated decisions. Difficult cases had no easy route to a second opinion.

What I built

• Case workflow. Referring hospitals submit cases, and each specimen is tracked through accessioning, digitisation, reporting and sign-out.
• Remote slide review. Pathologists open high-resolution whole-slide images in the browser from anywhere, using the WSI viewer I built for TPN.
• Consultations. Pathologists can bring a peer into a case to confirm a diagnosis before the report is released.
• Digital reporting. Structured reports go back to the referring clinician without paper or couriers.
• ML-assisted review. Machine-learning models help pathologists find regions of interest on digitised slides.
• Website redesign. A refreshed public site to support TPN's expansion to new hospitals.

Under the hood

Whole-slide images run to a gigabyte or more, so the platform never sends whole files to the browser. Slides are converted into tiled image pyramids and streamed on demand, which keeps viewing fast on ordinary hospital connections. Case data sits behind role-based access, so referring doctors, pathologists and administrators each see only what they should.

The result

Diagnoses that used to take weeks now come back within days. With working software in place, TPN could take its service to more hospitals across the country, and the same foundation later supported its pathology education hub, ALIS AC.