Advances in Leukemia detection and classification: A Systematic review of AI and image processing techniques

Area of Science:

• Hematology
• Medical Imaging
• Artificial Intelligence

Background:

• Leukemia presents complex diagnostic challenges globally, especially in resource-limited areas.
• Traditional diagnostic methods for leukemia face limitations in accuracy and speed.
• Advancements in AI and image processing offer potential solutions for improved leukemia detection.

Purpose of the Study:

• To systematically review the application of AI, particularly CNNs, in diagnosing four major leukemia types.
• To examine global epidemiological trends, risk factors, and healthcare disparities in leukemia.
• To assess the potential of AI in revolutionizing leukemia diagnostics and precision medicine.

Main Methods:

• Systematic literature review of over 25,000 articles from Scopus using PRISMA guidelines.
• Focused analysis on AI, specifically Convolutional Neural Networks (CNNs), for diagnosing ALL, AML, CLL, and CML.
• Examination of epidemiological data, risk factors, and healthcare access disparities.

Main Results:

• AI models, especially CNNs, show superior accuracy, speed, and reliability in leukemia diagnosis compared to conventional methods.
• Identified key risk factors including genetic syndromes, environmental toxins, radiation, and viral infections.
• Highlighted significant global disparities in leukemia incidence and outcomes influenced by socio-economic factors.

Conclusions:

• AI and image processing hold transformative potential for early leukemia detection, classification, and personalized treatment.
• Overcoming challenges like data variability, model scalability, and equitable AI access is crucial for global implementation.
• AI integration is vital for advancing precision medicine and improving leukemia patient outcomes worldwide.