Team
Hemo Hackers
“Non-invasive hematological monitoring using deep learning — detecting blood anemia through a standard smartphone camera, making clinical diagnostics accessible at every fingertip.”
Anemia Detection via Smartphone
Hemo Hackers developed a non-invasive hematological screening system that uses deep learning to detect blood anemia through the camera of a standard Android smartphone — no blood draw, no laboratory, no specialist required.
The system captures images of the patient’s conjunctiva (inner lower eyelid) and fingernail bed using the phone’s rear camera. A convolutional neural network trained on labelled hematological data classifies anaemia severity — mild, moderate, or severe — in real time, with accuracy benchmarked against standard haemoglobin testing.
The solution targets community health workers and sub-district health facilities that currently lack laboratory infrastructure. A health worker can screen a patient for anemia in under 60 seconds using a device that costs under $50 — turning any smartphone into a diagnostic tool.
Non-Invasive Screening
Conjunctiva and nail bed imaging — zero blood draw required.
Deep Learning Model
CNN trained on clinical hematological data. Sub-60 second inference.
Any Android Device
Works on sub-$50 smartphones. No specialist hardware or internet required.
Rural Deployment
Designed for community health workers and Upazila-level screening campaigns.
Problem Statement
Anemia affects approximately 40% of children and 30% of women in Bangladesh, contributing to maternal mortality, impaired cognitive development, and reduced workforce productivity. Yet diagnosis requires a complete blood count — a laboratory test inaccessible to the majority of the country’s rural population. Millions go undiagnosed until the condition is severe.
Event
- 9 May 2026DNA Hack For Health — Chittagong Divisional Round
- VenueChittagong Medical College, Chittagong
- Result1st Place — Chittagong Division Champion
