Using Deep Convolutional Neural Networks To Detect And Classify Indicators Of Leukemia In Blood Samples

The successful detection of leukemia depends on the correct interpretation of the lineage and morphology of monocytes — a type of white blood cells. The monocytes have been classified into four primary stages of differentiation. However, the stages cannot be distinguished unambiguously even by panels of trained experts. A number of attempts have been made to create tissue classifiers using deep learning that did not specifically address monocyte morphology. This research attempts to leverage whole-slide imaging and deep learning tools to create an automated high-accuracy method for cancerous monocytic cell detection.