Leukemia Detection Using Deep Learning: A DenseNet201-Based Approach

Leukemia is one of the most prevalent forms of blood cancer, primarily affecting white blood cells and posing critical diagnostic challenges. Conventional diagnostic approaches rely on manual microscopic examination, which is subjective, labor-intensive, and prone to errors. This study introduces a deep learning–based system for automated leukemia detection using microscopic images of blood smears. A pre-trained DenseNet201 model is applied through transfer learning, supported by preprocessing techniques such as normalization and augmentation to improve accuracy and reduce overfitting. The model is trained and validated on benchmark datasets, and its performance is measured using metrics including accuracy, precision, recall, and F1-score. Results demonstrate that the proposed system achieves high reliability, significantly reduces diagnostic latency, and provides a scalable framework suitable for integration into clinical workflows and mobile healthcare applications.