Advances in Automated Leukemia Detection Using Deep Analysis of WBC Scatterplot Patterns

Author: Nikul Vikaskumar Jayswal, Dr. Kirit J. Modi
Published Online: July 1, 2026
DOI: http://doi.org/10.63766/spujstmr.26.000087
Abstract
References

The development of computational hematology methods that use multidimensional white blood cell (WBC) scatterplots produced using modern hematology analyzers, automated leukemia diagnosis has experienced significant advancements. Based on light scatter, fluorescence and cellular distribution parameter variations, these scattering representations offer highly discriminative features to distinguish the leukemia subtypes with leukocyte abnormalities. Recent literature shows a distinct shift in traditional statistical and rule-based approaches to the use of advanced machine learning and deep learning models, such as convolutional neural networks, hybrid attention models, and transformers, which are effective in capturing complex spatial relationships in WBC clusters and have significantly higher diagnostic sensitivity and specificity. Simultaneous developments in the sphere of multimodal fusion models combining scatterplot functionality with clinical variables have only increased the reliability of diagnostic results. Additionally, explainable AI methods are also gradually being introduced to clarify important parts of scatter that affect the decision made by an algorithm and improve clinical interpretability. Even with these developments, there are still some remaining issues such as the lack of curated annotated datasets, inter-platform differences in analyzer results and large-scale clinical validation gaps. Other essential issues that are arising are the data privacy and safety of patients and this requires effective systems of encrypted data processing, encrypted model deployment and adherence to the laws of patient data security to ensure that sensitive patient information is safe. Such technical, clinical and ethical lapses need to be addressed in order to build reliable, generalizable and clinically implementable automated leukemia diagnostic systems.

Keywords: Automated leukemia diagnosis, White blood cell scatterplots, Machine Learning models, Deep learning models, Explainable artificial intelligence.
Download PDF Pages ( 201-207 ) Download Full Article (PDF)
←Previous Next →