Enhancing Acute Lymphoblastic Leukemia Classification with a Rapid and Effective CNN Model

Leukemias are malignant disorders of the hematopoietic stem cell compartment, characteristically associated with increased numbers of white cells in the bone marrow and/or peripheral blood. The course of leukemia may vary from a few days or weeks to many years, depending on the type (i.e., acute and chronic), which can be further classified into Acute lymphoblastic leukemia (ALL), Acute myeloid leukemia (AML), Chronic lymphocytic leukemia (CLL) and Chronic myeloid leukemia (CML). Given the high prevalence of ALL, a definite diagnosis necessitates intrusive, costly, and time-consuming tests. The nonspecific nature of the signs and symptoms associated with ALL often leads to misdiagnosis, further complicating the situation. In this study, a robust model is proposed that classifies the different subtypes of ALL. Our comprehensive diagnosis system achieved an accuracy of 97.85 % and demonstrated superior performance in comparison to other state-of-the-art approaches.