Automated Microscopic Malaria Diagnosis

In this project we will develop a computerized microscope based imaging system for field testing that will allow location and quantization of malarial parasites (trophozoites). The system will automatically scan a blood smear, locating the portions of the smear that are adequate for analysis and then counting the total number of red cells and the number of red cells containing parasites. The system will be capable of displaying digital images of all candidate parasites for confirmation by an operator if needed, and will allow relocation to objects located on the slides. The computer system will allow the operator to observe classes of located candidates (for example, high probability of being a parasite vs low probability) and confirm or reject objects. Total counts will be automatically updated during the editing process. High speed and high accuracy through sophisticated multi-spectral analysis are primary objectives of this project.