A Comprehensive Review of Segmentation and Classification of White Blood Cancer Cells in Bone Marrow Microscopic Images Using Deep Kronecker Neural Networks

Abstract

The detection and classification of white blood cancer cells, particularly leukemia, from bone marrow microscopic images are essential for early diagnosis and effective clinical decision-making. Traditional manual diagnostic approaches are labor-intensive, time-consuming, and prone to human error, creating a need for automated and reliable systems. Recent advancements in deep learning have significantly improved the accuracy of leukocyte segmentation and classification. This review focuses on state-of-the-art approaches, including convolutional neural networks (CNNs), U-Net-based segmentation models, transformer architectures, and emerging Deep Kronecker Neural Networks (DKNNs). CNN-based methods provide strong baseline performance, while U-Net variants enable precise segmentation of white blood cells in complex microscopic environments. Hybrid and attention-based models further enhance feature extraction and classification accuracy. Deep Kronecker Neural Networks offer a promising direction by efficiently modeling high-dimensional feature interactions with reduced computational complexity. Despite significant progress, challenges such as overlapping cells, staining variability, class imbalance, and limited datasets persist. This review presents a comprehensive analysis of existing methods, comparative evaluation, and future research directions aimed at developing robust, efficient, and clinically deployable leukemia detection systems.