Automatic Detection of Leukemia through Explainable AI-Based Machine Learning Approaches: Directional Review
Rida Arif, Shahzad Akbar, Sahar Gull, Qurat Ul Ain, Noor Ayesha · Auerbach Publications eBooks · 2025
Leukemia is a fatal blood cancer that affects people of all ages, accompanied by the excessive growth of white blood cells (WBCs) in the bone marrow. Hematologists usually recommend a bone marrow test to identify the presence of leukemia. Therefore, early-stage and precise detection of leukemia can save a human life. Recently, various traditional methods have been developed to diagnose leukemia that are time-consuming, unreliable, and need medical professionals in the diagnostic process. Consequently, leukemia is currently detectable and cured with different explainable artificial intelligence (XAI) techniques. This study provides a diversity of cutting-edge technologies that aid in the automatic identification of leucocytes. Moreover, various machine learning and deep learning methods for detecting leukemia at an early stage have been reviewed in this chapter. In this study, numerous leukemia detection methods have been reviewed after a detailed analysis of 44 pertinent articles, broadly divided into two main parts: (a) conventional machine learning–based leukemia detection and (b) deep learning–based leukemia detection. According to the findings, the most recent deep learning classifiers have been established to deliver accurate, swift, and reliable leukemia detection. Anyways, there are still required to publish more studies on the publicly accessible dataset for accurate identification and diagnosis of the other types of leukemia.