Standardised Transparent Orthopaedic Reporting and Modelling for AI (STORM-AI)-Guidelines for reporting artificial intelligence studies in orthopaedics from the ESSKA AI Working Group
Artikel i vetenskaplig tidskrift, 2026

Purpose The rapid growth of Artificial Intelligence (AI) in orthopaedic research has led to inconsistencies in study reporting, hindering evaluation and clinical translation. This initiative aimed to develop the STORM-AI (Standardised Transparent Orthopaedic Reporting and Modelling-AI) guidelines to enhance the transparency, completeness, and quality of reporting for AI studies in orthopaedics.Methods The ESSKA AI Working Group, a multinational and multidisciplinary team of experts, developed the STORM-AI guidelines through a multi-step consensus process. This involved a comprehensive review of existing AI reporting standards (e.g., CONSORT-AI, STARD-AI and TRIPOD), followed by iterative rounds of drafting, review, and refinement to incorporate orthopaedic-specific considerations.Results The consensus process resulted in the STORM-AI checklist and an accompanying Explanation and Elaboration (E&E) document. The guidelines provide specific reporting recommendations across all study sections, including study design, data characteristics, model development, performance metrics, ethical considerations and clinical workflow integration. Key areas of emphasis include rigorous validation, clear outcome definition, and error analysis within the orthopaedic context.Conclusion The STORM-AI guidelines provide a crucial framework for authors, reviewers, and journals to improve the evidence base for AI in orthopaedic care. Widespread adoption is anticipated to foster more robust, reproducible, and clinically valuable innovations, facilitating the responsible integration of AI into orthopaedics.Level of Evidence Level V.

guidelines

artificial intelligence

orthopaedics

reporting standards

machine learning

Författare

Felix C. Oettl

Universität Zürich

Balint Zsidai

Göteborgs universitet

Yinan Yu

Chalmers, Data- och informationsteknik, Funktionell programmering

James Pruneski

Tripler Army Medical Center

Thomas Tischer

Malteser Waldkrankenhaus Erlangen

Universität Rostock

Ayoosh Pareek

Hospital for Special Surgery - New York

Alberto Grassi

IRCCS Istituto Ortopedico Rizzoli, Bologna

Stefano Zaffagnini

IRCCS Istituto Ortopedico Rizzoli, Bologna

Michael T. Hirschmann

Universität Basel

Kantonsspital Baselland

Kristian Samuelsson

Göteborgs universitet

Journal of Experimental Orthopaedics

2197-1153 (eISSN)

Vol. 13 2 e70702

Ämneskategorier (SSIF 2025)

Ortopedi

DOI

10.1002/jeo2.70702

Mer information

Senast uppdaterat

2026-04-21