Automated Diagnosis of Acute Lymphoblastic Leukemia Leveraging EfficientNet and Fastai

This paper investigates a new way to improve the diagnosis of Acute Lymphoblastic Leukemia (ALL) using Convolutional Neural Networks (CNN). The method uses the EfficientNet-B7 model with the Fastai library to spot and group ALL from blood cell pictures. This tackles problems like not enough labeled data uneven classes, and differences in image data, which make accurate detection hard. The CNN model pulls out important features from the images on its making diagnosis much more accurate. The model's performance is evaluated on a dataset of blood smear images, achieving an accuracy of 99.69%, thereby demonstrating its effectiveness in distinguishing ALL-positive from ALL-negative samples. This research contributes to the advancement of automated diagnostic tools in healthcare, presenting a more efficient and reliable methodology for early and accurate ALL diagnosis.