Detecting small malignant B-cell populations using machine learning

hema.to’s AI model combines supervised and clustering approaches to detect small pathological B-cell populations in blood and bone marrow samples. Key features include:

• Efficient identification of populations as small as 3% of total cells
• Differentiation of six B-cell neoplasms: CLL, MCL, HCL, FL, MZL, and DLBCL
• Detection of previously unnoticed secondary B-NHL clones
• Optimization for a low false-negative rate, enabling fast, automated screening

The model offers systematic, rapid analysis to aid clinical B-cell neoplasm flow cytometry. View the full poster.