Deep learning enhances acute lymphoblastic leukemia diagnosis and classification using bone marrow images

Area of Science:

• Medical Imaging
• Artificial Intelligence in Healthcare
• Hematology

Background:

• Acute lymphoblastic leukemia (ALL) presents diagnostic challenges, especially in pediatric cases.
• Accurate and rapid diagnosis is crucial for effective ALL treatment.
• Bone marrow image analysis is a key component of ALL diagnostics.

Purpose of the Study:

• To review the role of deep learning (DL) in enhancing ALL diagnosis and classification.
• To analyze the performance of DL models in bone marrow image analysis for leukemia detection.
• To identify innovative DL techniques and their impact on diagnostic accuracy.

Main Methods:

• A systematic review of ten studies (2013-2023) on DL for ALL diagnosis.
• Analysis of Convolutional Neural Networks (CNNs) combined with techniques like Cat-Boosting, XG-Boosting, and Transfer Learning.
• Evaluation of DL models using performance metrics like accuracy.

Main Results:

• DL models, particularly CNNs, demonstrate high proficiency in detecting and classifying ALL from bone marrow images.
• Several DL models achieved accuracies of 99-100% in cancer cell classification.
• Models incorporating novel algorithms like Cat-Swarm Optimization showed superior classification performance.

Conclusions:

• Deep learning shows transformative potential for improving the efficiency and accuracy of ALL diagnostics.
• High accuracies achieved by DL models suggest a promising future for clinical integration.
• Further research is needed to address dataset limitations and refine DL models for optimal clinical use.