Document is current
Any future updates will be listed below
-
Authors
Antonissen, Noa https://orcid.org/0000-0001-8702-3640
Tryfonos, Olga
Houben, Ignas B.
Jacobs, Colin
de Rooij, Maarten
van Leeuwen, Kicky G.
- License Information
-
More Information
Article History
Received: 23 December 2024
Revised: 3 June 2025
Accepted: 17 June 2025
First Online: 24 July 2025
Compliance with ethical standards
:
: The scientific guarantor of this publication is Maarten de Rooij.
: The following authors of this manuscript declare relationships with the following companies: KvL is co-owner of Romion Health and Health AI Register and has received speaker fees from Siemens Healthineers and Bayer Pharmaceuticals. M.d.R. has an in-kind research agreement with Siemens Healthineers and has received speaker fees from Siemens Healthineers. C.J. receives research grants and royalties to the host institution from MeVis Medical Solutions, a public-private research grant with Philips Medical Systems, and a public-private research grant with Siemens Healthineers. C.J. received speaker fees from Canon Medical Systems and Johnson & Johnson. C.J. is a member of the Scientific Editorial Board of European Radiology (section: Imaging Informatics and Artificial Intelligence) and, as such, did not participate in the selection or review processes for this article. The other authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
: Gerjon Hannink kindly provided statistical advice for this manuscript.
: No human subjects were involved in this study.
: Institutional Review Board approval was not required because this study deals with public data and data provided by vendors.
: The collected data partially overlaps with the study by van Leeuwen et al (2021), which analysed peer-reviewed publications on 100 CE-certified radiological AI products and their supporting evidence. The current study builds upon this foundation by including additional AI products, updating the dataset to March 2023, and conducting a more comprehensive evaluation of peer-reviewed evidence. This overlap is clearly indicated in the Methods section, and the results are systematically compared to highlight trends and differences in study characteristics, efficacy levels, and overall evidence maturity [].
:
