Mainye, NyabutiMaranga, DawnAyako, RebeccahOchola, Lucy2024-03-162024-03-162022-06-16https://doi.org/10.31730/osf.io/jmtsqhttps://africarxiv.ubuntunet.net/handle/1/630https://doi.org/10.60763/africarxiv/586https://doi.org/10.60763/africarxiv/586https://doi.org/10.60763/africarxiv/586Introduction: Control of tropical diseases has for decades depended on diagnosis by microscopy as the gold standard method of detection. It, however, faces the drawbacks of low sensitivity, operator reliance on user expertise and experience essential to make an accurate diagnosis leading to variable results. Objective: The purpose of this study was to explore a method that could help improve the process of microscopy via machine learning models for the detection of intra- and intercellular parasites. Methods: A digital tool known as ‘cell explorer’ was developed to help in the detection and annotation of microscopic slide images taken from blood smears containing Leishmania donovani, Plasmodium falciparum and Trypanosoma brucei rhodesiense parasites. Advanced statistical modelling techniques were used including Convolutional Neural Networks, open-source image processing algorithms and clustering algorithms to detect cellular morphology of the parasites. Using Simple Linear Interactive Clustering, the cell explorer also functioned as a cell counter. Results: The neural network was able to immediately detect cellular morphology and identify the Leishmania and Trypanosome parasites as well as different stages of the Plasmodium parasite with an average accuracy of ̴ 95% . It was also able to accurately quantify the number of cells presented within each slide image. Conclusion: The cell explorer presents a fast and accurate computer-aided microscopy tool with the ability to detect cellular morphology, successfully identifying Leishmania donovani, Plasmodium berghei and Trypanosoma brucei rhodesiense parasites. This work highlights the research potential of machine learning models as disease diagnostic applications effective in improving the microscopy process.computer-aided microscopydeep learningDiagnosismachine learningparasitestropical diseasesMachine Learning Meets Microscopy: Cell Explorer Tool for a Diagnostic Laboratory