Malaria is transmitted through the bites of parasite-infected mosquitoes. With about 200 million cases worldwide, and about 400,000 deaths per year, malaria is a major global health problem. The standard method for malaria diagnosis is light microscopy of blood films. About 170 million blood films are examined every year for malaria, which involves manual counting of parasites. This is a tedious process, which depends heavily on the experience and skill of the microscopist. However, accurate parasite counts are essential to diagnosing malaria correctly, testing for drug-resistance, measuring drug-effectiveness, and classifying disease severity. Incorrect diagnostic decisions in the field can lead to unnecessary use of antibiotics or anti-malaria drugs, including side-effects and complications, or in some cases progression into potentially fatal severe malaria. To improve malaria diagnostics, the talk presents a fully-automated system for parasite detection and counting in blood films, which provides faster parasite counts compared to manual counting. The idea is to use inexpensive and portable smartphone technology to acquire blood film images in the field, with the smartphone’s camera attached to a microscope, and to run the automated diagnostic system software for these images on the phone. The system first detects individual red blood cells using image segmentation methods. In a second step, the system applies machine learning methods such as deep learning to classify between infected and uninfected cells. To learn the typical visual appearance of parasite-infected and uninfected red blood cells, the machine classifier has been trained on a large set of manually-annotated blood film images. The talk will discuss the image processing pipeline, challenges encountered, performance evaluation, and current field-testing. This will include the deep learning network used for cell classification and the execution of the network model on a smartphone.

Dr. Stefan Jaeger is an NIH Research Fellow at the U.S. National Library of Medicine (NLM), an institute within the National Institutes of Health (NIH). At NLM, he conducts research into image analysis and image informatics for clinical care and education. He leads a project team that develops computational screening methods for malaria, tuberculosis, and other diseases. Dr. Jaeger received his diploma in computer science from the University of Kaiserslautern and his PhD from the University of Freiburg, Germany. He is editorial board member of Quantitative Imaging in Medicine and Surgery and associate editor of Electronic Letters on Computer Vision and Image Analysis.