Artificial Intelligence Framework for Leukemia Detection and Classification Using Isurf-DLCNN from Blood Cell Image
Srinivas Bachu, Madasu Naga Venkata Akanksha · Apple Academic Press eBooks · 2025
It is anticipated that a total of 412,000 people would be diagnosed with leukemia worldwide, with acute lymphoblastic leukemia accounting for around 12% of all cases. Therefore, the early identification of leukemia has the potential to save the lives of millions of people. This article focuses mostly on the counting of blood cells and the diagnosis of leukemia via the use of deep learning processes. To begin, the images are pre-processed using median filters, and then the K-means clustering (KMC) method is used to segment the data. After that, the features are extracted by utilizing an improved speed up robust feature descriptor (ISURF), and then deep learning convolutional neural network (DLCNN) is used to classify the data based on these features. The proposed technique achieved an accuracy of 99% while maintaining a low level of complexity, and the proposed method achieved higher performance in comparison to the existing methods.