Detecting Acute Lymphocytic Leukemia in Individual Blood Cell Smear Images
Ruba Baluabid, Hadeel Alnasri, Rafaa Alowaybidi, Rawan Hafiz, Areej Alsini, Manal Alharbi · Engineering Technology & Applied Science Research · 2025
Acute Lymphocytic Leukemia (ALL) is a form of blood cancer that mainly affects lymphocytes and white blood cells. The severity of this cancer varies and progresses quickly, requiring immediate and intensive treatment and making a quick and accurate diagnosis essential. This study presents a diagnostic model for the diagnosis of ALL using deep learning. YOLOv8 achieved 95% accuracy when trained on the C-NMC dataset and 94% when trained on the ALL-IDB2 dataset while maintaining generalization. YOLOv8 outperformed other models such as SVM, ResNet-50, a hybrid model that integrates ResNet-50 with the SVM classifier, and DenseNet121. YOLOv8, with its strong architecture, can efficiently extract intricate patterns from medical imaging data and diagnose ALL. The proposed model can potentially reduce pathologist workloads and improve patient diagnosis. This research contributes to the field by providing a reliable tool for automated leukemia detection, paving the way for further advances in medical image analysis.