AI in hematology: Sepsis detection and differentiation

This study investigates how AI can be integrated into hematology workflows to overcome key challenges in sepsis diagnosis, including delayed detection, limited biomarker specificity, and nonspecific clinical presentations. By leveraging the HORIBA Yumizen H series analyzers together with the Generative Manifold Learning (GML) framework developed by GeodAIsics, Horiba aims to transform sepsis screening using only Complete Blood Count (CBC) data, enabling faster, more accurate, and accessible diagnostic support.