TN-Mammo: A Curated Multi-View Mammography Dataset with Consensus Radiologist Annotations for Breast Density Stratification in a Vietnamese Cohort
DOI:
https://doi.org/10.65405/t18s9k62Keywords:
Mammography dataset, breast density classification, BI-RADS, multi-view imaging, radiologist consensus, Vietnamese cohort, medical image annotation, AI-ready data curation.Abstract
Digital mammography is still the principal method for population-level screening, and early detection of breast cancers is still essential for successful clinical management. Modern computational methods, especially those that use deep learning architectures, have demonstrated significant promise in improving radiological evaluation; however, the representativeness and annotation fidelity of underlying training repositories inherently limit their effectiveness. Current public mammography collections sometimes have issues with multi-view completeness, annotation processes, or demographic diversity, which restricts generalizability across diverse clinical contexts. This research has represented the TN-Mammo, a carefully selected multi-view mammography dataset from a Vietnamese patient cohort, to fill in these gaps. Each participant contributed anatomically paired left and right breast pictures to the library, which includes bilateral craniocaudal (CC) and mediolateral oblique (MLO) projections for 676 individuals. Two board-certified radiologists separately assessed breast density categorization, a crucial indicator for cancer risk stratification, using a double-blind methodology. Consensus adjudication was used to determine final labels. To ensure clinical interoperability, density assignments adhere to the defined four-tier BI-RADS framework (categories A–D). The acquisition process, annotation technique, inter-observer agreement measures, and baseline statistical characterizations of the dataset are all described in this paper. This research hope to assist the development of fair, population-sensitive AI systems for breast cancer screening in underrepresented areas and enable repeatable research in density-aware computer-aided diagnosis by making TN-Mammo publically available via PhysioNet.
Downloads
References
[1] Heath, M., et al. (2000). The Digital Database for Screening Mammography. IWDM.
[2] Lee, R. S., et al. (2017). A curated mammography dataset for computer-aided detection and diagnosis. Scientific Data, 4, 170177.
[3] Nguyen, H. Q., et al. (2021). VinDr-Mammo: A large-scale benchmark dataset for computer-aided diagnosis in breast cancer screening. arXiv preprint arXiv:2110.13130.
[4] Dembrower, K., et al. (2020). EMBED: A multi-ethnic mammography dataset for breast density assessment. Medical Physics, 47(12), 6235–6245.
[5] American College of Radiology. (2013). ACR BI-RADS® Atlas, Breast Imaging Reporting and Data System (5th ed.). Reston, VA.
[6] Nguyen, B., Le, C., Vu, L., Nguyen, Q., Pham, H. H., Vu, P. A., ... & PhysioBank, P. (2025). TN-Mammo: A Multi-view Mammography Dataset for Breast Density Classification. https://doi.org/10.13026/34kz-bk76
[7] Ben Dalla, L. O. F., Medeni, T. D., Medeni, I. T., & Ulubay, M. (2025). Enhancing Healthcare Efficiency at Almasara Hospital: Distributed Data Analysis and Patient Risk Management. Economy: Strategy and Practice, 19(4), 54–72. https://doi.org/10.51176/1997-9967-2024-4-54-72
[8] Dalla, L. O. F. B. (2020). The Influence of hospital management framework by the usage of Electronic healthcare record to avoid risk management (Department of Communicable Diseases at Misurata Teaching Hospital: Case study). EHRM, 20(4), 22–52. https://doi.org/20.51176/1954-9923-2020-4-22-52
[9] Gergerli, B., Çelebi, F. V., Rahebi, J., & Şen, B. (2023). An Approach Using in Communication Network Apply in Healthcare System Based on the Deep Learning Autoencoder Classification Optimization Metaheuristic Method. Wireless Personal Communications, 1-24.
[10] Dalla, L. O. F. B. (2020). Dorsal Hand Vein (DHV) Verification in Terms of Deep Convolutional Neural Networks with the Linkage of Visualizing Intermediate Layer Activations Detection. International Journal of Engineering and Modern Technology E-ISSN 2504-8848 P-ISSN 2695-2149 Vol 6 No 2 (2020). www.iiardpub.org
[11] Dalla, L. O. F. B. (2020). Convolutional Neural Network Baseline Model Building for Person Re-Identification. International Journal of Engineering and Modern Technology E-ISSN 2504-8848 P-ISSN 2695-2149 Vol. 6 No. 3 2020 www.iiardpub.org
[12] Karal, Ö., & Dalla, L. O. F. B. Lung Nodule Characterization in CT Scans Using Hybrid 3D Attention U-Net Segmentation and Transfer Learning-Based Classification Approach. Comprehensive Journal of Science, Volume (10), Issue (37), (NOV. 2025) Special issue for the Third International Conference on Science and Technology, www.sicst.ly, SICST2025, ISSN: 3014-6266, Reply: 6266-3014
[13] Çakır, M., Degirmenci, A., & Karal, O. (2022, February). Exploring the behavioural factors of cervical cancer using ANOVA and machine learning techniques. In International Conference on Science, Engineering Management and Information Technology (pp. 249-260). Cham: Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-40395-8_18
[14] Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and psychological measurement, 20(1), 37-46.https://doi.org/10.1177/001316446002000104
[15] Cohen, J. (1968). Weighted kappa: Nominal scale agreement provision for scaled disagreement or partial credit. Psychological Bulletin, 70(4), 213–220. https://doi.org/10.1037/h0026256
[16] McGrath, J. E., & Meyer, G. J. (2006). When effect sizes disagree: The case of r and d. Psychological Methods, 11(4), 386-401. https://psycnet.apa.org/buy/2006-22258-004
[17] Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33(1), 159-174. https://doi.org/10.2307/2529310
[18] Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. api.taylorfrancis.com
[19] Lachin, J. M. (2008). Sample size evaluation for a multiply studied case in the placebo-controlled Diabetes Control and Complications Trial (DCCT). Statistics in Medicine, 27(14), 2541-2558. First published: 20 Sept (2007). https://doi.org/10.1002/sim.3057
[20] Biau, D. J., Kernéis, S., & Porcher, R. (2008). Statistics in brief: The importance of sample size in the planning and interpretation of medical research. Clinical Orthopaedics and Related Research, 466(9), 2282-2288. https://doi.org/10.1007/s11999-008-0346-9
[21] Shannon, C. E. (1948). A mathematical theory of communication. The Bell System Technical Journal, 27(3), 379-423. https://doi.org/ 10.1002/j.1538-7305.1948.tb01338.x
[22] Hüllermeier, E., & Waegeman, W. (2021). Aleatoric and epistemic uncertainty in machine learning: An introduction to concepts and methods. Machine Learning, 110(3), 457-506. https://doi.org/ 10.1007/s10994-021-05946-3
[23] Sinecen, M., Cinar, M., Karal, O., Engin, M., Atesci, Y. Z., Makinaci, M., & Cakmak, B. (2009, May). Diagnosis of Prostat Cancer using Artificial Neural Networks. In 2009 14th National Biomedical Engineering Meeting (pp. 1-3). IEEE https://doi.org/10.1109/BIYOMUT.2009.5130296
[24] M. Yumus, M. Apaydin, A. Degirmenci, and O. Karal, “Missing data imputation using machine learning based methods to improve HCC survival prediction,” 2020 28th Signal Processing and Communications Applications Conference, SIU 2020 - Proceedings, Oct. 2020, doi: 10.1109/SIU49456.2020.9302222.
[25] Dalla, L. O. F. B. (2020). Convolutional Neural Network Baseline Model Building for Person Re-Identification. International Journal of Engineering and Modern Technology E-ISSN 2504-8848 P-ISSN 2695-2149 Vol. 6 No. 3 2020 www.iiardpub.org
[26] Dalla, L. O. F. B. (2020). The Influence of hospital management framework by the usage of Electronic healthcare record to avoid risk management (Department of Communicable Diseases at Misurata Teaching Hospital: Case study). E-ISSN 2876-9948 P-ISSN 2453-2149 Vol. 3 No. 1 2020 www.iiardpub.org
[27] Yalman, Y., Uyanık, T., Atlı, İ., Tan, A., Bayındır, K. Ç., Karal, Ö., ... & Guerrero, J. M. (2022). Prediction of voltage sag relative location with data-driven algorithms in distribution grid. Energies, 15(18), 6641.Energies 2022, 15(18), 6641; https://doi.org/10.3390/en15186641
[28] Nguyen, H. T., Nguyen, H. Q., Pham, H. H., Lam, K., Le, L. T., Dao, M., & Vu, V. (2023). VinDr-Mammo: A large-scale benchmark dataset for computer-aided diagnosis in full-field digital mammography. Scientific Data, 10(1), 277. https://www.nature.com/articles/s41597-023-02100-7
[29] Nguyen, B., Le, C., Vu, L., Nguyen, Q., Pham, H. H., Vu, P. A., ... & PhysioBank, P. (2025). TN-Mammo: A Multi-view Mammography Dataset for Breast Density Classification. https://doi.org/10.13026/1kx0-xc60
[30] Nguyen, B., Le, C., Vu, L., Nguyen, Q., Pham, H., Vu, P. A., Huynh, T., Tien Dung, C., Diep Tuong, N., & Hong, B. (2025). TN-Mammo: A Multi-view Mammography Dataset for Breast Density Classification (version 1.0.0). PhysioNet. RRID:SCR_007345. https://doi.org/10.13026/1kx0-xc60
[31] Goldberger, A., Amaral, L., Glass, L., Hausdorff, J., Ivanov, P. C., Mark, R., ... & Stanley, H. E. (2000). PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals. Circulation . 101 (23), pp. e215–e220. RRID:SCR_007345.
[32] Lim, Y. X., Lim, Z. L., Ho, P. J., & Li, J. (2022). Breast cancer in Asia: incidence, mortality, early detection, mammography programs, and risk-based screening initiatives. Cancers, 14(17), 4218; https://doi.org/10.3390/cancers14174218
[33] Nguyen, H. G. (2024). Identification of Asymptomatic Vertebral Fracture Using Artificial Intelligence Methods (Doctoral dissertation, University of Technology Sydney (Australia)).
[34] Nickson, C., Velentzis, L. S., Mann, G. B., Grogan, P., Bateson, D., & Canfell, K. (2025). Risk-adjusted breast screening: an Australian perspective and considerations for the Western Pacific. The Lancet Regional Health–Western Pacific, 57. https://doi.org/PIIS2666-6065(25)00057-4/fulltext
[35] Hill, H., Kearns, B., Duffy, S., Pashayan, N., Sasieni, P., & Offman, J. (2024). Estimating the cost-effectiveness of risk stratified breast cancer screening in the UK. https://doi.org/10.1186/s12919-024-00306-0PDF
[36] Tun, H. M., Rahman, H. A., Naing, L., & Malik, O. A. (2025). Artificial intelligence utilization in cancer screening program across ASEAN: a scoping review. BMC cancer, 25(1), 703. https://doi.org/10.1186/s12885-025-14026-x
[40] Ben Dalla, L, O, F. (2021). Literature review (LR) on the powerful of Research methodology processes life cycle. In 2021 The Powerful of Research Methodology Processes Life Cycle Conference (TPRMPLCC) (pp. 1-10). IEEE. https://doi.org/10.16543/TPRMPLCC 50717.2020.92876580
[41] Muttaqi, M., Degirmenci, A., & Karal, O. (2022, September). US accent recognition using machine learning methods. In 2022 Innovations in Intelligent Systems and Applications Conference (ASYU) (pp. 1-6). IEEE. https://doi.org/10.1109/ASYU56188.2022.9925265
[42] Dulkadir, S. E. Z. G. İ. N., Tecimer, H. U., Parlaktürk, F., Altındal, Ş., & Karal, Ö. M. E. R. (2020). The effect of radiation on the forward and reverse bias current–voltage (I–V) characteristics of Au/(Bi4Ti3O12/SiO2)/n-Si (MFIS) structures. Journal of Materials Science: Materials in Electronics, 31(15), 12514-12521.https://doi.org/10.1007/s10854-020-03801-0
[43] Arık, D. T., Karal, Ö., & Şahin, A. B. (2020). A Comparative Study of Artificial Neural Networks and Naïve Bayes Techniques for the Classification of Radar Targets. Bitlis Eren Üniversitesi Fen Bilimleri Dergisi, 9(4), 1779-1788.https://doi.org/10.17798/bitlisfen.676973
[44] Uysal, Z., Kalkancı, G., İmren, T., Değirmenci, A., Karal, Ö., & Çankaya, İ. (2016). A Heart Rate Monıtoring Applicatıon Using Wireless Sensor Network System Based on Bluetooth Wıth Matlab GUI. Int. J. Eng. Sci, 6, 2862.International Journal of Engineering Science and Computing, August 2016 , http://ijesc.org/
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Comprehensive Journal of Science
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
