Medical Imaging Lab
- DICOM networking
- CT reconstruction
- ultrasound imaging
- image processing
- Hounsfield Units
- window and level
- PACS architecture
- modality workflows
- medical image formats
- reconstruction algorithms
Research & Labs
Learning by Building
This section documents technical investigations, research notes, prototypes and engineering laboratories. Not everything here is a commercial product — some projects exist purely to understand the science and engineering more deeply.
Clinical AI Lab
- CAD systems
- segmentation
- classification
- object detection
- explainable AI
- dataset shift
- model calibration
- uncertainty
- medical AI validation
- multimodal clinical systems
Speech & Language AI Lab
- Whisper
- medical speech recognition
- speaker diarization
- LLM fine-tuning
- RAG
- QLoRA
- quantization
- local inference
- GPU optimisation
Biomedical Systems Lab
- ESP32
- Arduino
- sensors
- MQTT
- physiological signals
- IoT
- medical instrumentation
- robotics
Agricultural Intelligence Lab
- satellite imagery
- vegetation indices
- disease detection
- yield prediction
- model fusion
- edge AI
- precision agriculture
Open Questions
Current Research Questions
- 01
How well do medical AI models trained elsewhere generalise to African populations?
- 02
What medical imaging infrastructure architecture makes sense for low-resource environments?
- 03
How can useful AI systems run on limited GPU hardware?
- 04
How can radiology workflows be improved using AI without removing clinicians from the decision loop?
- 05
How should African healthcare datasets be governed, anonymised and shared responsibly?
- 06
What combination of cloud, edge and on-premise computing best suits healthcare environments with inconsistent connectivity?
Related writing lives in Engineering Writing. Product context for Zyemed Labs is in the Zyemed case study.