Medical Imaging · Digital Health · AI
Zyemed — Medical Imaging Infrastructure
Modular medical imaging infrastructure spanning PACS, RIS, reporting intelligence and healthcare AI research for African clinical environments.
Role: Founder / Biomedical Software Engineer
- Python
- FastAPI
- Node.js
- React
- PostgreSQL
- Orthanc
- OHIF
- Docker
- Linux
- S3-compatible Object Storage
- Redis
- DICOMweb
Overview
Zyemed is a healthcare technology platform focused on medical imaging infrastructure, clinical workflows and healthcare AI — designed for African operational realities and currently serving clients in production.
The Problem
Diagnostic imaging depends on far more than acquiring an X-ray, CT, MRI or ultrasound image.
Behind every study is an ecosystem involving patient information, modality communication, study storage, image retrieval, clinical viewing, reporting, workflow management, security, long-term archiving and interoperability.
Many healthcare facilities still work with fragmented workflows and infrastructure that can be expensive or difficult to adapt to local operational realities.
Context
Imaging centres and hospitals need systems that remain interoperable with standard modalities while supporting cloud and on-premise deployment models — including environments where connectivity, specialist access and computing resources may be constrained.
My Role
Founder and biomedical software engineer responsible for architecture, backend services, DICOM/DICOMweb integration paths, product direction across Lumina (PACS), Orbis (RIS), Sonus (reporting intelligence) and Labs (AI & data).
Architecture
Rendering diagram…
Engineering Decisions
- Use standards-first imaging protocols (DICOM / DICOMweb) rather than proprietary lock-in
- Keep Orthanc and OHIF as proven building blocks where they accelerate interoperability
- Separate PACS, RIS, reporting and research concerns into modular products
- Design for multi-tenancy, auditability and both cloud and local deployment
Challenges
- multi-tenancy and patient privacy
- network reliability and image transfer economics
- DICOM interoperability across vendors
- latency and clinical usability
- security, observability and disaster recovery
Solutions
Modular service boundaries, standards-based interfaces, object storage for study objects, careful tenant isolation, and progressive product layers (archive → workflow → reporting → AI research).
Current Status
Production — Zyemed is a working medical imaging product with clients. Lumina (PACS) and Orbis (RIS) are in active use; Sonus and Labs continue as advancing product and research layers.
Lessons
Healthcare infrastructure is a systems problem. Architecture, interoperability and operational constraints matter as much as feature velocity.