Emerging Trends in Leukemia Diagnosis: Innovations, Challenges, and Future Perspectives

Leukemia is a blood and bone marrow disorder caused by the uncontrolled proliferation of white blood cells, which overcrowd normal cells and disrupt bone marrow function. It is broadly categorized into four main types: Acute Lymphoblastic Leukemia (ALL), Chronic Lymphocytic Leukemia (CLL), Acute Myeloid Leukemia (AML), and Chronic Myeloid Leukemia (CML). Further subtypes are classified based on leukemia cell characteristics, genetic mutations, and other factors. Accurate prognosis is essential for guiding treatment decisions. Traditional diagnostic techniques include cell morphology analysis and cytometry-based methods, such as microarray analysis. However, morphology-based tests are prone to diagnostic errors, while cytometry techniques are costly. To overcome these challenges, Machine Learning (ML) and Deep Learning (DL) approaches have gained prominence in blood image classification for leukemia detection. Researchers have explored ML-based classification models, while deep learning techniques leverage pre-trained models—such as those trained on ImageNet or medical imaging datasets—that are fine-tuned with leukemia-specific data. This paper explores machine learning-based solutions and deep learning architectures for leukemia classification, discussing key challenges, current advancements, and future research directions.