article · Discover Artificial Intelligence
Acute myeloid leukaemia is an aggressive blood cancer requiring precise identification of abnormal white blood cells for diagnosis. Distinguishing between different leukaemia-related cell types is difficult because various white blood cells look very similar visually. Deep learning models, specifically YOLOv12, Inception-ResNet-v2, and ResNet50, were evaluated for multiclass cell classification alongside cell-level and nucleus-level segmentation techniques using Hue-channel extraction and Otsu thresholding. Preserving full cell morphology rather than focusing only on the nucleus substantially enhanced classification accuracy. The combination of YOLOv12 with Otsu-based cell segmentation delivered the top performance, reaching a test accuracy of 99.3 per cent. In comparison, Inception-ResNet-v2 achieved 97.6 per cent accuracy, and ResNet50 reached 93.87 per cent. Pairing modern transfer learning architectures with effective segmentation provides clear discrimination among challenging white blood cell categories, including monocytes and basophils.
Accurate diagnosis of acute myeloid leukaemia relies on differentiating visually similar white blood cells, which can be difficult using conventional microscopy analysis. Demonstrating that deep learning models can achieve up to 99.3 per cent accuracy across challenging cell types highlights how automated image segmentation and classification can support reliable, highly sensitive identification of malignant blood cells.
This technology could support diagnostic software for haematology laboratories and clinical pathologists seeking automated white blood cell classification. While the models demonstrate high accuracy on test data, the findings represent early-stage applied research evaluated in an experimental setting, meaning clinical validation and integration into digital pathology workflows are still required before real-world adoption.
AI-generated from the published abstract. Always read the original work before citing.
Abstract Introduction Acute Myeloid Leukemia (AML) is an aggressive hematological malignancy whose diagnosis relies heavily on the accurate identification of abnormal white blood cells. However, multiclass classification of AML-related cell types remains challenging due to the high visual similarity among different leukocyte categories. Methods This study investigates the classification of multiclass AML-related cells using three deep learning models: ResNet50, Inception-ResNet-v2, and YOLOv12. Two segmentation strategies based on cell-level and nucleus-level representations were evaluated using Hue-channel extraction and Otsu thresholding to enhance morphological feature extraction prior to classification. Results Experimental results show that YOLOv12 combined with Otsu-based cell segmentation achieved the highest test accuracy of 99.3%. Inception-ResNet-v2 achieved a test accuracy of 97.6% using Hue-based cell segmentation, while ResNet50 achieved 93.87% using nucleus-based Otsu segmentation. The findings demonstrate that preserving complete cellular morphology significantly improves classification performance. Conclusion The results highlight the effectiveness of combining segmentation techniques with modern transfer learning architectures for AML-related cell classification. In particular, Otsu-based cell segmentation and YOLOv12 provided the most discriminative representation of cellular morphology, enabling highly accurate classification of challenging cell types such as basophils and monocytes.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1007/s44163-026-01833-9
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
