Explainable AI for Hematological Diagnostics: Integrating Image Processing and LIME for Leukemia Detection

This study introduces an innovative system for diagnosing leukemia, leveraging the power of artificial intelligence (AI) combined with deep learning and Explainable AI (XAI) to build confidence in AI-driven medical tools. The system processes images of blood cells to precisely identify and classify different types of leukemia, such as Acute Lymphoblastic Leukemia (ALL), Acute Myeloid Leukemia (AML), Chronic Lymphocytic Leukemia (CLL), and Chronic Myeloid Leukemia (CML). To ensure transparency, the model employs XAI methods, delivering clear and interpretable predictions along with counterfactual explanations that explore alternative diagnostic possibilities. This not only strengthens the system’s reliability but also promotes accountability in medical decision-making. Beyond diagnosis, the system provides customized treatment recommendations based on the patient’s unique medical background, effectively connecting accurate diagnostics with practical, actionable steps. By focusing on precision, transparency, and a patient-first approach, this solution not only enhances the detection of leukemia but also builds trust and understanding among healthcare providers regarding the use of AI in critical medical scenarios.