Related Experiment Video
Updated: Jan 8, 2026
Evaluation of a Point-of-Care Testing Analyzer for Measuring Peripheral Blood Leukocytes
Published on: March 22, 2022
Evaluation of a Semi-Supervised AI Model (ASUS Blade) for Peripheral Blood Film Leukocyte Classification
Bingwen Eugene Fan1,2,3,4, David Tao Yi Chen5, Chiew Yan Lee1
1Department of Haematology, Tan Tock Seng Hospital, Singapore.
Background:
The peripheral blood film (PBF) analysis traditionally relies on manual microscopy (MM), a labour-intensive method with inter-observer variability. This study evaluates Blade (a semi-supervised AI model) and CellaVision DM9600 (commercial benchmark) against MM in automated leukocyte classification.
Methods:
PBFs from 168 patients were prepared using automated staining and scanned digitally. Blade, trained on 185 412 cells (75 435 labelled, 109 977 unlabelled) via ResNet34 and RetinaNet architectures, underwent pseudo-labelling and AdamW optimisation. Performance was evaluated on 1675 cells against MM using the concordance correlation coefficient (CCC), Bland-Altman analysis, Deming/Passing-Bablok regression and diagnostic accuracy measures across nine leukocyte subtypes.
Results:
When evaluated individually against MM, both systems showed high agreement. Blade achieved excellent correlation for common cells (neutrophils: ccc = 0.988; lymphocytes: ccc = 0.985; eosinophil: ccc = 0.953) and comparable results to CellaVision for monocytes (ccc = 0.852 vs. 0.847) and basophils (ccc = 0.762 vs. 0.794). Blade performed better for metamyelocytes (ccc = 0.905 vs. 0.756) and showed higher sensitivity for monocytes (75% vs. 63%) and myelocytes (87% vs. 74%). Regression analysis showed slopes close to 1.0 for most cell types, with Blade displaying narrower Limits of Agreement in Bland-Altman analysis. Both systems achieved 100% sensitivity for blasts and reactive lymphocytes. Overall macro-averaged performance was comparable between Blade (sensitivity 89.2%, specificity 96.3%) and CellaVision (86.3% and 96.7%).
Conclusion:
Blade and CellaVision demonstrated strong concordance with MM, validating their clinical utility. Blade's semi-supervised learning confers marginal advantages in rare cell detection and stability, highlighting AI's potential to enhance diagnostic accuracy. While both systems reduce labour and variability, Blade's performance has potential for integration into haematology workflows. Future validation in diverse cohorts is recommended.
Insights
This study shows that the AI model Blade and the CellaVision DM9600 system closely match manual microscopy (MM) for automated leukocyte classification in peripheral blood film analysis. Blade demonstrates potential for enhancing diagnostic accuracy and integrating into hematology workflows.
Area of Science:
- Hematology
- Artificial Intelligence
- Medical Diagnostics
Background:
- Peripheral blood film (PBF) analysis traditionally uses manual microscopy (MM), which is labor-intensive and prone to inter-observer variability.
- Automated methods are being developed to improve efficiency and consistency in PBF analysis.
Purpose of the Study:
- To evaluate the performance of a semi-supervised AI model, Blade, and a commercial system, CellaVision DM9600, against manual microscopy for automated leukocyte classification.
- To assess the diagnostic accuracy and concordance of these automated systems across various leukocyte subtypes.
Main Methods:
- Digital scanning of PBFs from 168 patients.
- Training the Blade AI model using ResNet34 and RetinaNet architectures on a large dataset of cells.
- Evaluating Blade and CellaVision DM9600 against manual microscopy using concordance correlation coefficient (CCC), Bland-Altman analysis, and regression techniques.
Main Results:
- Both Blade and CellaVision showed high agreement with manual microscopy for leukocyte classification.
- Blade achieved excellent correlation for common leukocytes (e.g., neutrophils, lymphocytes) and demonstrated superior performance for metamyelocytes compared to CellaVision.
- Both systems achieved 100% sensitivity for blasts and reactive lymphocytes, with comparable overall macro-averaged performance.
Conclusions:
- Blade and CellaVision DM9600 are clinically useful automated systems for leukocyte classification, showing strong concordance with manual microscopy.
- Blade's semi-supervised approach offers potential advantages in detecting rare cells and improving diagnostic stability.
- AI-driven tools like Blade have the potential to enhance diagnostic accuracy and streamline hematology workflows.
