Authors
Thomas Kau, Julia Furtner-Srajer, Gabriel Adelsmayr, Gerlig Widmann, Christian Neumann, Matthias Perkonigg, Ramona Woitek, Kristina Lang, Giovanni Rubeis, Markus Pedevilla, Michael Gruber, Helmut Prosch, Antonius Schuster, Nikolaus Forgo, Janne Cadamuro, Barbara Prainsack, Michael H Fuchsjäger, Georg Langs, Martin Michael Uggowitzer
Published in
RoFo : Fortschritte auf dem Gebiete der Rontgenstrahlen und der Nuklearmedizin. Oct 27, 2026. Epub Oct 27, 2026.
Abstract
This position paper was written at the request of the executive board of the Austrian Radiological Society. As a co-opted member of the executive board of the Austrian Radiological Society for digitalization and artificial intelligence (AI), the corresponding author was tasked with assembling an interdisciplinary group of authors. The authors were selected on the basis of their scientific and technical expertise, their position at clinical and university facilities, and their involvement in national and international professional societies and working groups related to AI in radiology. Various radiology subspecialties, the major university health care regions in Austria, and health facilities on various care levels are represented. Experts from the fields of technology, law, and ethics were also included. The referenced literature is based on targeted research without a systematic meta-analysis. The individual chapters were written by topic-specific subgroups and the chapters were then combined by the first author to create this manuscript. The Delphi method was not used. The expert consensus was reviewed by the executive board of the Austrian Radiological Society and approved as the official position of the society after revision.
· Safe implementation of commercial AI requires standardized local validation, clearly defined responsibilities, continuous performance monitoring, version control, and established procedures for error reporting, revalidation, and deactivation.. · AI performance cannot be established by regulatory approval or technical benchmarks alone but must be demonstrated under local clinical conditions.. · Human responsibility remains essential despite increasing automation.. · AI is shifting the role of radiology from primarily image interpretation toward an increasingly data-integrative, prognostic, and interdisciplinary discipline.. · The value of AI should be assessed by patient-relevant outcomes and effects on healthcare delivery rather than by technical performance or workforce reduction alone..
· Kau T, Furtner-Srajer J, Adelsmayr G et al. Artificial Intelligence in Radiology: A Position Paper of the Austrian Radiological Society (ÖRG). Rofo 2026; DOI 10.1055/a-2965-2624.
PMID:
42945971
Bibliographic data and abstract were imported from PubMed on 28 Oct 2026.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 36
- Comments 0