Leukemia Blood Cancer Detection

Abstract

Automatically generated identification of cancers of the leukocytes, which include Leukemia, is a tricky biomedical research problem. Many sorts of laboratory procedures that take hours are used to manually detect leukemia. In this paper, we have suggested a method for detecting leukemia that is both quick and accurate. The proposed methodology intends to detect leukocyte cancer early, reduce cases of false positives, and improve the learning methodology of the system. Here we have implemented 3 different pre- trained deep learning models namely Inceptionv3, Xception and Resnet50 for identifying the cancer cells presence. And all the 3 models performed well with the accuracy of 80%, 92.88 % and 98.15% respectively.