Cloud Health Data Platform on Azure

Moving eye-diagnostic data to the cloud: a health data platform for a medical technology company

MADIS engineers help build an event-driven, cloud-native platform on Microsoft Azure that brings all ophthalmic measurement data, images and videos together in one place.

At a glance

Cloud-native

event-driven, serverless architecture

FHIR + DICOM

healthcare interoperability standards

Multi-client

serves web and mobile applications

Summary: Measurement data from eye-diagnostic devices used to be stored on local computers in clinics. For a German medical technology company, MADIS helps build a cloud-native Health Data Platform on Azure that centralises measurements, images and metadata, streams videos online, serves web and mobile applications, and lays the foundation for AI-supported analytics.

The client

The client is a German medical technology company that develops and manufactures eye-examination devices and equipment used in eye surgery. Its goal is to digitalise and connect healthcare ecosystems through interoperable digital products.

The challenge

The client’s earlier data management software had an older architecture: measurement data was stored and used on computers in the clinics themselves. That made it hard to give several applications access to the same data, and limited what could be built on top of it.

The new platform had to work in a regulated medical environment, follow healthcare interoperability standards and handle large files such as diagnostic videos.

• Healthcare standards: devices record data in DICOM (patient, physician and measurement metadata, images, videos, PDFs), while web applications work better with FHIR.
• Large media: videos had to be viewable online, not only downloadable.
• Scale of delivery: several backend teams in different countries working in parallel within a scaled agile framework.

Our approach

Event-driven, serverless backend

The platform is built as an event-driven, cloud-native application on Microsoft Azure, using serverless services so that developers can focus on business logic rather than infrastructure.

Standards-based data handling

Measurement data, images and metadata are stored and managed using DICOM and FHIR on Azure Health Data Services, with adaptive streaming for online video playback.

Infrastructure as Code and DevOps

Infrastructure is managed with Terraform. MADIS engineers take part in development, testing, DevOps and infrastructure tasks, and in writing the technical documentation.

Scaled agile delivery

Work is organised in product increments under the Scaled Agile Framework (SAFe), with backend teams from several countries working on the same platform.

The solution

The result is a scalable health data platform delivered as SaaS, where all measurement data – including images, videos and metadata – can be found in one place. It provides centralised access for several client applications, both web and mobile, and enables interconnected devices and applications to support clinical workflows remotely.

Technologies: Microsoft Azure · Azure Health Data Services · FHIR · DICOM · TypeScript · Node.js · Serverless · Terraform

Results

• One place for all data: measurements, images, videos and metadata are centralised instead of being kept on local clinic computers.
• Ready for many applications: centralised access serves several web and mobile clients.
• Foundation for AI: the platform lays the groundwork for an AI-supported analytics component, which the client has started developing.

Facing a similar challenge?

Planning a move from on-premises software to a cloud-native platform? Talk to our software and cloud team.