Age and mammographic morphology in tumor classification: a comparison using multilayer perceptron neural networks

Authors

Keywords:

Breast Neoplasms; Age Groups; Women's Health; Comparative Study; Neural Networks

Abstract

Introduction: Clinical and imaging differences according to age in breast tumors justify comparing the findings by using neural networks to estimate their diagnostic relevance.

Objective: To compare the importance of mammographic features and age in the histopathological classification of breast nodules.

Materials and Methods: An analytical observational study based on 831 complete records from the Mammographic Mass Data dataset (447 benign and 384 malignant) was performed. Multilayer perceptron neural networks were developed for the ≤49- and ≥50-year age groups. Accuracy; sensitivity; specificity; predictive values; area under the curve; and the importance of age, shape, margins and density were evaluated.

Results: The training error was 13.2% in the ≤49-year group and 20.3% in the ≥50-year group; the testing error was 17.6% and 16.8%, respectively. Accuracy was 0.85 and 0.88; sensitivity, 0.82 and 0.85; specificity, 0.87 and 0.90; and AUC, 0.864 and 0.837. In the ≤49-year group, shape had the greatest importance (0.470; 100%), followed by margins (0.399; 85%). In the ≥50-year group, age predominated (0.359; 100%), followed by shape (0.335; 93.3%).

Conclusions: The neural network showed that tumor severity-related factors vary by age. In patients <50 years, shape and margins are more relevant, whereas in those ≥50 years, age predominates.

Downloads

Download data is not yet available.

References

1. Cai Y, Dai F, Ye Y, Qian J. The global burden of breast cancer among women of reproductive age: a comprehensive analysis. Sci Rep [Internet]. 2025;15(1):9347. Disponible en: http://doi.org/10.1038/s41598-025-93883-9

2. Mohapatra SK, Das PK, Nayak RB, Mishra A, Nayak B. Diagnostic accuracy of mammography in characterizing breast masses using the 5th edition of BI-RADS: a retrospective study. Cancer Res Stat Treat [Internet]. 2022;5(1):52–8. Disponible en: http://doi.org/10.4103/crst.crst_224_21

3. Ren T, Gao Z, Yang L, Cheng W, Luo X. Development of a nomogram for predicting malignancy in BI-RADS 4 breast lesions using contrast-enhanced ultrasound and shear wave elastography parameters. Sci Rep [Internet]. 2025;15(1):1356. Disponible en: http://doi.org/10.1038/s41598-025-85862-x

4. Mustafa W, Mustafa O. ACR breast density: relationship with age and its impact on mammographic and ultrasound findings. Alq J Med App Sci [Internet]. 2024;1:847–54. Disponible en: http://doi.org/10.54361/ajmas.247356

5. Al-Balas H, Al-Balas M, Herz Allah S, Qudsi A, Ramadan MS, Al Tawil G, et al. The association of mammographic breast density and breast cancer biology among breast cancer women in Jordan. Arch Breast Cancer [Internet]. 2024;11(4). Disponible en: http://doi.org/10.32768/abc.2024114378-384

6. Gomede E. Unlocking complexities: the role of nonlinear relationships and hidden layers in neural networks [Internet]. EE UU: medium; 2024. Disponible en: https://medium.com/the-modern-scientist/unlocking-complexities-the-role-of-nonlinear-relationships-and-hidden-layers-in-neural-networks-fca9631fd2fd

7. Elter M. Mammographic Mass [Internet]. EE UU: UCI Machine Learning Repository; 2007. Disponible en: http://doi.org/10.24432/C53K6Z

8. Engmann NJ, Scott C, Jensen MR, Winham SJ, Ma L, Brandt KR, et al. Longitudinal changes in volumetric breast density in healthy women across the menopausal transition. Cancer Epidemiol Biomarkers Prev [Internet]. 2019;28(8):1324–30. Disponible en: http://doi.org/10.1158/1055-9965.EPI-18-1375

9. Brooks JD, Christensen RAG, Sung JS, Pike MC, Orlow I, Bernstein JL, et al. MRI background parenchymal enhancement, breast density and breast cancer risk factors: a cross-sectional study. NPJ Breast Cancer [Internet]. 2022;8(1):97. Disponible en: http://doi.org/10.1038/s41523-022-00458-2

10. Moffat RE, Eichholzer M, Myrick ME, Schmid SM, Raggi A, de Geyter C, et al. Menopausal state in breast cancer: how reliable is the data? Clin Breast Cancer [Internet]. 2011;11(6):390–4. Disponible en: http://doi.org/10.1016/j.clbc.2011.07.002

11. Valdez BJ, Grumley M, Chang SC, Keller JK, Grumley JG, Orozco JIJ. Contemporary trends in breast cancer in females under the age of fifty: an NCDB study. Surg Oncol Insight [Internet]. 2024;1(2):100049. Disponible en: http://doi.org/10.1016/j.soi.2024.100049

12. Molefi T, Mabonga L, Hull R, Mwazha A, Sebitloane M, Dlamini Z. The histomorphology to molecular transition: exploring the genomic landscape of poorly differentiated epithelial endometrial cancers. Cells [Internet]. 2025;14(5). Disponible en: http://doi.org/10.3390/cells14050382

13. Sabit H, Adel A, Abdelfattah MM, Ramadan RM, Nazih M, Abdel-Ghany S, et al. The role of tumor microenvironment and immune cell crosstalk in triple-negative breast cancer: emerging therapeutic opportunities. Cancer Lett [Internet]. 2025;628:217865. Disponible en: http://doi.org/10.1016/j.canlet.2025.217865

14. Conti M, Morciano F, Amodeo S, Gori E, Romanucci G, Belli P, et al. Special types of breast cancer: clinical behavior and radiological appearance. J Imaging [Internet]. 2024;10(8):182. Disponible en: http://doi.org/10.3390/jimaging10080182

15. Jiang X, Shen H, Shang X, Fang J, Lu Y, Lu Y, et al. Recent advances in the aging microenvironment of breast cancer. Cancers (Basel) [Internet]. 2022;14(20):4990. Disponible en: http://doi.org/10.3390/cancers14204990

16. Testa U, Castelli G, Pelosi E. Breast cancer: a molecularly heterogeneous disease needing subtype-specific treatments. Med Sci (Basel) [Internet]. 2020;8(1):18. Disponible en: http://doi.org/10.3390/medsci8010018

17. Makki J. Diversity of breast carcinoma: histological subtypes and clinical relevance. Clin Med Insights Pathol [Internet]. 2015;8:CPath.S31563. Disponible en: http://doi.org/10.4137/CPath.S31563

18. Lloyd R, Pirikahu S, Walter J, Cadby G, Warrington N, Perera D, et al. The prospective association between early life growth and breast density in young adult women. Cancers (Basel) [Internet]. 2024;16(13):2418. Disponible en: http://doi.org/10.3390/cancers16132418

19. Kerlikowske K, Cook AJ, Buist DSM, Cummings SR, Vachon C, Vacek P, et al. Breast cancer risk by breast density, menopause, and postmenopausal hormone therapy use. J Clin Oncol [Internet]. 2010;28(24):3830–7. Disponible en: http://doi.org/10.1200/JCO.2009.26.4770

Published

2026-11-16

How to Cite

1.
Guevara Tirado A. Age and mammographic morphology in tumor classification: a comparison using multilayer perceptron neural networks. Rev haban cienc méd [Internet]. 2026 Nov. 16 [cited 2026 Nov. 17];25:e6160. Available from: https://revhabanera.sld.cu/index.php/rhab/article/view/6160

Issue

Section

Clinical and pathological sciences