Malaria Detection: AI-Powered Blood Analysis on a Smartphone
CodeTherapy + Dr. Mulugeta's Lab | June 2, 2025 | 5 min read

CodeTherapy partnered with Dr. Mulugeta's research group in Ethiopia to develop an on-device machine learning model for complete blood count analysis using affordable, smartphone-mounted microscopes. The collaborative effort leverages the widespread ubiquity of mid-tier consumer smartphones to bypass the need for multi-thousand-dollar laboratory hardware.
The co-development team built an image-processing pipeline capable of identifying and counting blood cells in live microscope feeds. By optimizing the neural network models for native execution on simple mobile microprocessors, the tool functions entirely offline, ensuring that isolated clinics with poor cellular connectivity retain robust diagnostic capabilities.
This portable tool helps rural health workers conduct hematology tests without needing centralized laboratory infrastructure, potentially transforming malaria screening in remote communities across East Africa. Early clinical tests show dramatic decreases in diagnostic time, reducing wait-times from hours to under five minutes.
Related Projects

Mammography AI for Early Detection
Deep-learning screening models for breast cancer optimized for compressed, low-contrast clinical mammograms in resource-constrained regional clinics.
Read more
AI-Powered Laparoscopy
An end-to-end laparoscope embedded with low-cost IoT sensors and real-time onboard computer vision, built in coordination with Senegalese biomedical engineers.
Read moreLet’s co-create pathways for diagnostic equity.
We partner with medical practitioners, academic institutions, and funding groups to refine, validate, and deploy open-source healthcare technologies where they are needed most.
