Retrospective evaluation of a CE-marked AI system, including 1,017,208 mammography screening examinations

Sep 23, 2026

Uploaded by McGucket

Retrospective evaluation of a CE-marked artificial intelligence system assesses performance in identifying breast cancer on screening mammograms. Using screening data from 2004–2021 across ten breast centers in BreastScreen Norway, radiologists scored each case under an independent double-reading protocol while the AI generated NT and SN risk scores. Operating points N70 and P3 defined negative and positive cases, enabling analysis of misclassification and cancer detection behavior. Results covered 1,017,208 examinations and informed workload reduction and sensitivity implications.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

Introduction

Background and rationale

Screening benefits and harms

Role of AI in mammographic screening

Need for validation before implementation

Materials and methods

Results

Conclusion

Key Points

Hovda et al. European Radiology (2025) 35:5685–5694 [https://doi.org/10.1007/s00330-025-1](https://doi.org/10.1007/s00330-025-1)1521-4 BREAST Open Access Retrospective evaluation of a CE-marked AI system, including 1,017,208 mammography screening examinations Tone Hovda 1, Marthe Larsen2, Marie Burns Bergan2, Jonas Gjesvik2, Lars A. Akslen3,4 and Solveig Hofvind2,5* Abstract Objectives To retrospectively evaluate the performance of a CE-marked AI system for identifying breast cancer on screening mammograms. Evidence from large retrospective studies is crucial for planning prospective studies and to further ensure safe implementation. Materials and methods We used data from screening examinations performed from 2004 to 2021 at ten breast centers in BreastScreen Norway. In the standard independent double reading setting, each radiologist scored each breast from 1 (negative) to 5 (high probability of cancer) . The AI system assigned each examination an NT and an SN score; the NT score aimed to classify examinations as negative with minimal misclassiﬁcation while the SN score aimed to classify examinations as positive with high conﬁdence. N70 was deﬁned as being among the 70% with the lowest NT score and P3 was deﬁned as being among the 3% with the highest SN score. Results A total of 1,017,208 screening examinations were included in the study sample. At N70, 1 . 8%(107/5977) of the screen-detected and 34 . 5%(625/1812) of the interval cancers were deﬁned as negative. Using P3 to deﬁne cases as positive, 81.5%(4871/5977) of the screen-detected and 19.0%(344/1812) of the interval cancers were deﬁned as positive. Among the screen-detected cancers in N70, 11 . 2%(12/107) had an interpretation score > 2 by both radiologists. Conclusion The AI system performed well according to identifying negative cases and cancer cases. Thus, the AI system can be used to reduce workload for the radiologists and potentially increase the sensitivity of mammography. Key Points Question Results from large mammography screening samples not used in training AI algorithms are important to consider when planning prospective studies and implementation. Findings More than 80% of the screening-detected cancers were classiﬁed as positive by AI when considering 3% of the examinations with the highest AI risk score as positive. Clinical relevance A lack of radiologists is a challenge in mammographic screening. Our ﬁndings support other studies that suggest the use of AI to reduce screen-reading workload. Keywords Breast cancer, Mammography, Artiﬁcial intelligence, Screening *Correspondence: Solveig Hofvind [sshh@kreftregisteret. no](sshh@kreftregisteret. no) Full list of author information is available at the end of the article © The Author(s) 2025 . Open Access This article is licensed under a Creative Commons Attribution 4 . 0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit [http://creativecommons.org/](http://creativecommons.org/)[ ](http://creativecommons.org/)[licenses/by/4.0/](licenses/by/4.0/.)[.](licenses/by/4.0/.) Graphical Abstract Introduction Breast cancer is the most frequent cancer and among the leading causes of cancer deaths among women worldwide [1]. Mammographic screening is recommended as secondary prevention to reduce breast cancer m

FAQs

What was the study objective in evaluating the CE-marked AI system?

To retrospectively evaluate how the CE-marked AI system performs in identifying breast cancer on screening mammograms, supporting safe and effective planning for prospective use.

How were negative and positive cases defined using the AI scores?

The AI assigned an NT score and an SN score. N70 selected cases among the 70% with the lowest NT score (defined as negative), and P3 selected cases among the 3% with the highest SN score (defined as positive).

What do the results indicate about AI use in screening?

At the operating points, the AI classified more than 80% of screen-detected cancers as positive when using the top 3% of AI risk scores. The authors conclude the AI performed well for identifying both negative cases and cancer cases, potentially reducing radiologist workload and increasing sensitivity.

Sep 23, 2026

Uploaded by McGucket