Deep learning prediction of mammographic breast density using screening data.

Publisher: Nature Publishing Group Country of Publication: England NLM ID: 101563288 Publication Model: Electronic Cited Medium: Internet ISSN: 2045-2322 (Electronic) Linking ISSN: 20452322 NLM ISO Abbreviation: Sci Rep Subsets: MEDLINE

Original Publication: London : Nature Publishing Group, copyright 2011-

Deep Learning* ; Breast Density* ; Mammography*/methods ; Breast Neoplasms*/diagnostic imaging ; Breast*/diagnostic imaging; Humans ; Female ; Retrospective Studies ; Middle Aged ; Adult ; Aged ; Early Detection of Cancer/methods ; Radiographic Image Interpretation, Computer-Assisted/methods

Competing Interests: Declarations. Competing interests: The authors declare no competing interests. Ethical approval: Due to the retrospective nature of the study, the Medical Ethics Committee of Taizhou Cancer Hospital waived the need of obtaining informed consent (IRB-2023-002-IIT), and the study was conducted in compliance with the principles of the Declaration of Helsinki and its contemporary amendments, as well as Good Clinical Practice.
This study investigated a series of deep learning (DL) models for the objective assessment of four categories of mammographic breast density (e.g., fatty, scattered, heterogeneously dense, and extremely dense). A retrospective analysis was conducted using data collected from Taizhou Cancer Hospital over a period from January 2015 to December 2020. The dataset included mammograms from 9,621 women, totaling 57,282 images. The dataset was divided into training, validation, and test sets at a ratio of 7:2:1. Four DL models were employed, with Average Precision (AP) served as the primary evaluation metric. Additionally, the diagnostic performance of the DL models was compared with that of radiologists. Finally, we conducted validation of our model using an external test set. Among the DL models studied, InceptionV3 performed best, with AP values of 0.895 for almost entirely fatty, 0.857 for scattered fibroglandular tissue, 0.953 for heterogeneously dense, and 0.952 for extremely dense categories. The InceptionV3 model outperformed radiologists in accuracy and consistency. While radiologists surpassed the InceptionV3 model in fatty and scattered categories, their accuracy dropped significantly in heterogeneously and extremely dense categories. Nevertheless, our study demonstrated that DL can serve as a valuate tool in assisting radiologists with the objective quantification of breast density.
(© 2025. The Author(s).)

Radiol Artif Intell. 2020 Nov 04;3(1):e200015. (PMID: 33937850)
J Med Imaging (Bellingham). 2019 Jul;6(3):031405. (PMID: 30746393)
Breast Cancer Res Treat. 2017 Apr;162(3):541-548. (PMID: 28161786)
Breast. 2012 Aug;21(4):503-6. (PMID: 22285387)
J Clin Oncol. 2004 Jul 15;22(14):2842-8. (PMID: 15254051)
CA Cancer J Clin. 2022 Jan;72(1):7-33. (PMID: 35020204)
Radiology. 2020 Feb;294(2):265-272. (PMID: 31845842)
Breast Cancer Res. 2022 Feb 20;24(1):14. (PMID: 35184757)
JAMA Netw Open. 2021 Dec 1;4(12):e2139161. (PMID: 34940866)
AJR Am J Roentgenol. 2012 Mar;198(3):W292-5. (PMID: 22358028)
Cancer. 2020 Nov 1;126(21):4687-4696. (PMID: 32767699)
JAMA. 2012 Apr 4;307(13):1394-404. (PMID: 22474203)
Acad Radiol. 2021 Apr;28(4):475-480. (PMID: 32089465)
Radiology. 2019 Jan;290(1):52-58. (PMID: 30325282)
Breast Cancer Res. 2007;9(6):217. (PMID: 18190724)
Med Phys. 2020 Jan;47(1):110-118. (PMID: 31667873)
CA Cancer J Clin. 2021 May;71(3):209-249. (PMID: 33538338)
J Natl Cancer Inst. 2003 Jan 1;95(1):30-7. (PMID: 12509398)
J Natl Cancer Inst. 1995 May 3;87(9):670-5. (PMID: 7752271)
Semin Cancer Biol. 2021 Jul;72:214-225. (PMID: 32531273)
N Engl J Med. 2019 Nov 28;381(22):2091-2102. (PMID: 31774954)
Acad Radiol. 2019 Apr;26(4):544-549. (PMID: 30072292)
JAMA. 2008 May 14;299(18):2151-63. (PMID: 18477782)
CA Cancer J Clin. 2022 Nov;72(6):524-541. (PMID: 36190501)
JAMA Oncol. 2023 Jun 1;9(6):808-814. (PMID: 37103922)
Ann Intern Med. 2016 Oct 4;165(7):457-464. (PMID: 27428568)
Radiol Artif Intell. 2022 Mar 16;4(2):e210199. (PMID: 35391766)
Radiology. 2021 Nov;301(2):283-292. (PMID: 34402665)
Breast Cancer Res. 2016 May 03;18(1):45. (PMID: 27142210)
J Natl Cancer Inst. 2005 Sep 21;97(18):1366-76. (PMID: 16174858)
Eur Radiol. 2021 Sep;31(9):7192-7201. (PMID: 33738595)

2023C04039 "Pioneer" and "Leading Goose" R&D Program of Zhejiang; CSJRZC2021JJSJ001 Research Program of National Health Commission Capacity Building and Continuing Education Center

Keywords: Automated breast density quantification; Breast cancer risk; Deep learning; Mammography

Date Created: 20250404 Date Completed: 20250515 Latest Revision: 20250515

20250519

PMC11971370

10.1038/s41598-025-95275-5

40185813