Development and multicenter validation of an AI-driven detection system for DLBCL cells in bone marrow smears
6 affiliations
- 1Department of Hematology-Oncology, Hematologic Oncology Intelligent Diagnosis and Treatment Engineering Research Center of Chongqing Education Commission of China, Chongqing University Cancer Hospital
- 2Department of Hematology, Southwest Hospital of Third Military Medical University (Army Medical University)
- 3Department of Haematology, Daping Hospital, Third Military Medical University
- 4Department of Laboratory Medicine, Chongqing Center for Clinical Laboratory, Chongqing Academy of Medical Sciences, Chongqing General Hospital
- 5School of Medicine, Chongqing University
- 6College of Computer Science, Chongqing University
Abstract
The presence of diffuse large B-cell lymphoma (DLBCL) cells in bone marrow (BM) smears is recognised as a key morphological basis for diagnosing BM involvement by DLBCL. BM involvement directly affects disease staging, treatment strategies, and prognosis assessment. Nevertheless, traditional manual identification of DLBCL cells in BM smears under a microscope is time-consuming, subject to observer variability, and lacks standardization between hospitals. In this study, we developed an artificial intelligence (AI) detection system (DLBCL-CQnet) based on deep learning, using BM smears from 117 patients with newly diagnosed DLBCL across four centers. DLBCL-CQnet comprises a region of interest (ROI) classification model based on MobileNetV2, and a cell detection and classification model (CDC-MOD) based on YOLOv8. Performance was evaluated using accuracy, precision, recall, F1-score, and mean average precision at an intersection over union threshold of 0.5 (mAP50). MobileNetV2 achieved the best ROI classification performance (accuracy 0.925, precision 0.922, F1-score 0.919), while YOLOv8 yielded a classification accuracy of 0.890 and an mAP50 of 0.934 for DLBCL cell detection. The system achieved an ROI classification accuracy of 0.925, a cell classification accuracy of 0.890, and a mean average precision (mAP50) of 0.934 at an intersection-over-union (IoU) threshold of 0.5. The optimal cutoff for identifying BM infiltration by DLBCL cells was 2.12%, with a specificity of 0.941 and positive predictive value (PPV) of 0.923, demonstrating high diagnostic value. The recognition speed significantly outperforms manual assessment, substantially reducing time consumption. DLBCL-CQnet enables automated, standardized identification of DLBCL cells in BM smears across institutions, serving as a reliable auxiliary diagnostic tool for assessing BM involvement by DLBCL. Its high specificity and operational efficiency streamline diagnostic workflows, especially in resource-constrained environments.
Related
- Development and validation of a laboratory-developed test on Panther Fusion(R) system for detection of macrolide resistance in Bordetella pertussis
- Development and internal validation of a clinical risk score for early delirium detection in older emergency department patients
- Development and multicenter validation of a machine learning model for predicting in-hospital postoperative cardiac arrest in elderly surgical patients: the critical role of intraoperative hemodynamic disturbances and blood transfusion.
- Development and validation of an artificial intelligence model for basal cell carcinoma detection and section completeness assessment in Mohs micrographic surgery in an Asian cohort
- Development and external validation of machine learning models for predicting mortality in patients with lung cancer: a multicenter retrospective cohort study
BibTeX
@article{ran2026development,
title = {{Development and multicenter validation of an AI-driven detection system for DLBCL cells in bone marrow smears}},
author = {Longrong Ran and Yu Peng and Xuelian Wu and Yuanyou Yao and Zheng Li and Shuyue Chen and Wei Zhang and Xiaojun Xia and Yakun Zhang and Mingyu Zhao and Shuang Chen and Can Lin and Lian Li and Liangmei Li and Jun Li and Yang Li and Xing Wu and Zailin Yang and Yao Liu},
journal = {Scientific Reports},
year = {2026},
month = jul,
doi = {10.1038/s41598-026-59908-7},
url = {https://doi.org/10.1038/s41598-026-59908-7},
note = {CC BY},
}
RIS
TY - JOUR
TI - Development and multicenter validation of an AI-driven detection system for DLBCL cells in bone marrow smears
AU - Longrong Ran
AU - Yu Peng
AU - Xuelian Wu
AU - Yuanyou Yao
AU - Zheng Li
AU - Shuyue Chen
AU - Wei Zhang
AU - Xiaojun Xia
AU - Yakun Zhang
AU - Mingyu Zhao
AU - Shuang Chen
AU - Can Lin
AU - Lian Li
AU - Liangmei Li
AU - Jun Li
AU - Yang Li
AU - Xing Wu
AU - Zailin Yang
AU - Yao Liu
PY - 2026
DA - 2026/07/01
JO - Scientific Reports
DO - 10.1038/s41598-026-59908-7
UR - https://doi.org/10.1038/s41598-026-59908-7
AB - The presence of diffuse large B-cell lymphoma (DLBCL) cells in bone marrow (BM) smears is recognised as a key morphological basis for diagnosing BM involvement by DLBCL. BM involvement directly affects disease staging, treatment strategies, and prognosis assessment. Nevertheless, traditional manual identification of DLBCL cells in BM smears under a microscope is time-consuming, subject to observer variability, and lacks standardization between hospitals. In this study, we developed an artificial intelligence (AI) detection system (DLBCL-CQnet) based on deep learning, using BM smears from 117 patients with newly diagnosed DLBCL across four centers. DLBCL-CQnet comprises a region of interest (ROI) classification model based on MobileNetV2, and a cell detection and classification model (CDC-MOD) based on YOLOv8. Performance was evaluated using accuracy, precision, recall, F1-score, and mean average precision at an intersection over union threshold of 0.5 (mAP50). MobileNetV2 achieved the best ROI classification performance (accuracy 0.925, precision 0.922, F1-score 0.919), while YOLOv8 yielded a classification accuracy of 0.890 and an mAP50 of 0.934 for DLBCL cell detection. The system achieved an ROI classification accuracy of 0.925, a cell classification accuracy of 0.890, and a mean average precision (mAP50) of 0.934 at an intersection-over-union (IoU) threshold of 0.5. The optimal cutoff for identifying BM infiltration by DLBCL cells was 2.12%, with a specificity of 0.941 and positive predictive value (PPV) of 0.923, demonstrating high diagnostic value. The recognition speed significantly outperforms manual assessment, substantially reducing time consumption. DLBCL-CQnet enables automated, standardized identification of DLBCL cells in BM smears across institutions, serving as a reliable auxiliary diagnostic tool for assessing BM involvement by DLBCL. Its high specificity and operational efficiency streamline diagnostic workflows, especially in resource-constrained environments.
SN - 2045-2322
ER -