[ECP 2026] Diagnostic Performance of the miLab™ BCM for Morphological Detection of Blasts and Immature Granulocytes in Peripheral Blood

miLab™ BCM showed close agreement with expert manual smear review and high NPV for blast and IG flags, which supports its potential as a first-line CBC and morphology screening tool.

ECP 2026 (European Congress of Pathology)

September 12, 2026

C.Y. Bae, et al.

[ECP 2026] Diagnostic Performance of the miLab™ BCM for Morphological Detection of Blasts and Immature Granulocytes in Peripheral Blood

Unauthorized downloading or distribution of this poster is strictly prohibited.

Supplementary Video | Reviewing AI-Flagged Cells in the miLab™ Viewer

Abstract

This study, run with Asan Medical Center, compared the miLab™ BCM with expert manual peripheral blood smear review in 361 samples. It showed close agreement across the 6-part differential and sensitive flagging of immature granulocytes (100%) and blasts (88.9%) at clinically significant thresholds.

Key Highlights

Single-center study at Asan Medical Center (Seoul): 361 evaluable peripheral blood smear cases, compared with a certified medical technologist who has over 10 years of morphology experience.
miLab™ BCM runs an automated CBC and AI-based WBC classification from 5 μL of whole blood in one cartridge-based run.
Strong agreement across the 6-part differential: Passing-Bablok slopes near unity for neutrophils (1.01), lymphocytes (0.97), eosinophils (0.99) and IG (0.94).
IG flag: 100% sensitivity, 100% NPV, ROC AUC 0.996. Blast flag: 88.9% sensitivity, 99.4% NPV, ROC AUC 0.935.
Results reflect the AI model's standalone output before reviewer confirmation, and flagged cells can be confirmed in the on-screen gallery.

Download Brochure Request Demo