Comparative Analysis of the Performance of Automated Digital Cell Morphology Analyzers for Leukocyte Differentiation in Hematologic Malignancies: Mindray MC-80 Versus West Medical Vision Hema

Introduction

The use of artificial intelligence in hematology laboratories has improved the diagnostic evaluation of peripheral blood cells. The aim of this study is to compare the performance of two automated digital cell morphology analyzers, the Mindray MC-80 and the West Medical Vision Hema Pro, with manual microscopy, the gold standard, for leukocyte differentiation in patients with hematologic malignancies and infections.

Methods

Peripheral blood smears from 75 patients were analyzed, including cases of acute lymphoblastic leukemia (ALL, 4), chronic lymphocytic leukemia (CLL, 20), acute myeloid leukemia (AML, 20), chronic myeloid leukemia (CML, 5), other lymphoproliferative disorders (LPD, 20), and infections (6). The agreement between microscopy, Vision Hema, and MC-80 was assessed by Bland–Altman analysis for eight leukocyte populations (neutrophils, lymphocytes, monocytes, eosinophils, basophils, band cells, myelocytes, and metamyelocytes).

Results

Vision Hema demonstrated better agreement with manual microscopy for eight normally expected leukocyte populations (neutrophils, lymphocytes, monocytes, eosinophils, basophils, band cells, myelocytes, and metamyelocytes), whereas MC-80 exhibited greater biases, particularly in lymphocytes, basophils, and immature granulocytes. For pathologic cells, VH significantly overestimated blasts, while MC-80 classified them more accurately, showing better agreement with manual microscopy in acute leukemias. Additionally, MC-80 showed potential clinical value in detecting abnormal lymphocytes and promyelocytes, which may be relevant for hematologic malignancies.

Conclusion

Vision Hema provides more reliable classification of normally expected leukocyte populations, while MC-80 shows advantages in detecting abnormal cells, particularly in hematologic malignancies.