Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Nov 7;7(1):69.
doi: 10.1186/s41747-023-00384-3.

Deep learning performance for detection and classification of microcalcifications on mammography

Filippo Pesapane  1 , Chiara Trentin  2 , Federica Ferrari  2 , Giulia Signorelli  2 , Priyan Tantrige  3 , Marta Montesano  2 , Crispino Cicala  4 , Roberto Virgoli  4 , Silvia D'Acquisto  4 , Luca Nicosia  2 , Daniela Origgi  5 , Enrico Cassano  2
Affiliations

Deep learning performance for detection and classification of microcalcifications on mammography

Filippo Pesapane et al. Eur Radiol Exp. .

Abstract

Background: Breast cancer screening through mammography is crucial for early detection, yet the demand for mammography services surpasses the capacity of radiologists. Artificial intelligence (AI) can assist in evaluating microcalcifications on mammography. We developed and tested an AI model for localizing and characterizing microcalcifications.

Methods: Three expert radiologists annotated a dataset of mammograms using histology-based ground truth. The dataset was partitioned for training, validation, and testing. Three neural networks (AlexNet, ResNet18, and ResNet34) were trained and evaluated using specific metrics including receiver operating characteristics area under the curve (AUC), sensitivity, and specificity. The reported metrics were computed on the test set (10% of the whole dataset).

Results: The dataset included 1,000 patients aged 21-73 years and 1,986 mammograms (180 density A, 220 density B, 380 density C, and 220 density D), with 389 malignant and 611 benign groups of microcalcifications. AlexNet achieved the best performance with 0.98 sensitivity, 0.89 specificity of, and 0.98 AUC for microcalcifications detection and 0.85 sensitivity, 0.89 specificity, and 0.94 AUC of for microcalcifications classification. For microcalcifications detection, ResNet18 and ResNet34 achieved 0.96 and 0.97 sensitivity, 0.91 and 0.90 specificity and 0.98 and 0.98 AUC, retrospectively. For microcalcifications classification, ResNet18 and ResNet34 exhibited 0.75 and 0.84 sensitivity, 0.85 and 0.84 specificity, and 0.88 and 0.92 AUC, respectively.

Conclusions: The developed AI models accurately detect and characterize microcalcifications on mammography.

Relevance statement: AI-based systems have the potential to assist radiologists in interpreting microcalcifications on mammograms. The study highlights the importance of developing reliable deep learning models possibly applied to breast cancer screening.

Key points: • A novel AI tool was developed and tested to aid radiologists in the interpretation of mammography by accurately detecting and characterizing microcalcifications. • Three neural networks (AlexNet, ResNet18, and ResNet34) were trained, validated, and tested using an annotated dataset of 1,000 patients and 1,986 mammograms. • The AI tool demonstrated high accuracy in detecting/localizing and characterizing microcalcifications on mammography, highlighting the potential of AI-based systems to assist radiologists in the interpretation of mammograms.

Keywords: Artificial intelligence; Machine learning; Mammography; Microcalcifications; Neural networks (computer).

PubMed Disclaimer

Conflict of interest statement

FP and DO are members of the European Radiology Experimental Scientific Editorial Board. They have not taken part in the review or selection process of this article. Laife Reply provided funding to the IEO-European Institute of Oncology as part of a collaboration agreement which included the author’s research activities for the present study.

Figures

Fig. 1
Fig. 1
Study flowchart illustrating the workflow of the study, highlighting the key steps involved in the analysis of mammographic images for microcalcifications classification. The study follows a 5-point framework, encompassing data collection, anonymization, annotation, analysis of annotated data, and network training and evaluation. The flowchart provides a visual representation of the interplay between these phases and the various patient subdivisions, including training, validation, and testing. Numbers of patches are indicated to convey the distribution of microcalcifications and their nature (benign/malignant)
Fig. 2
Fig. 2
Example of annotations of suspicious microcalcifications in craniocaudal (a) and medio-lateral (b) mammograms performed by radiologists using a special application for tagging (X-RAIS, see the “Methods”)
Fig. 3
Fig. 3
Example of heatmaps generated by Task 1 AlexNet predictions (a, c), compared with radiologists’ annotations (b, d). Colour changes based on the probability of the predictions. The colour scale visually represents the probability of microcalcifications in the area; it ranges from blue to red, which are 0% and 100%, respectively
Fig. 4
Fig. 4
Examples of benign microcalcifications (a–e) incorrectly classified cases as malignant and malignant microcalcifications (f–l) incorrectly classified as benign
Fig. 5
Fig. 5
Examples of areas with microcalcifications (a–e) incorrectly classified cases as areas without microcalcifications and areas without microcalcifications (f–l) incorrectly classified as areas with microcalcifications
See this image and copyright information in PMC

References

    1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–249. doi: 10.3322/caac.21660. - DOI - PubMed
    1. Fuller MS, Lee CI, Elmore JG. Breast cancer screening: an evidence-based update. Med Clin North Am. 2015;99:451–468. doi: 10.1016/j.mcna.2015.01.002. - DOI - PMC - PubMed
    1. Bevers TB, Helvie M, Bonaccio E et al (2018) Breast Cancer Screening and Diagnosis, Version 3.2018, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 16(11):1362–1389. 10.6004/jnccn.2018.0083 - PubMed
    1. Lehman CD, Topol EJ. Readiness for mammography and artificial intelligence. Lancet. 1867;2021:398. doi: 10.1016/S0140-6736(21)02484-3. - DOI - PubMed
    1. Sechopoulos I, Teuwen J, Mann R. Artificial intelligence for breast cancer detection in mammography and digital breast tomosynthesis: state of the art. Semin Cancer Biol. 2021;72:214–225. doi: 10.1016/j.semcancer.2020.06.002. - DOI - PubMed

Publication types

LinkOut - more resources