Graf, Robert http://orcid.org/0000-0001-6656-3680
Platzek, Paul-Sören
Riedel, Evamaria Olga
Kim, Su Hwan
Lenhart, Nicolas
Ramschütz, Constanze
Paprottka, Karolin Johanna
Kertels, Olivia Ruriko
Möller, Hendrik Kristian
Atad, Matan
Bülow, Robin
Werner, Nicole
Völzke, Henry
Schmidt, Carsten Oliver
Wiestler, Benedikt
Paetzold, Johannes C.
Rueckert, Daniel
Kirschke, Jan Stefan
Funding for this research was provided by:
HORIZON EUROPE European Research Council (101045128—iBack-epic—ERC2021-COG)
Article History
Received: 10 May 2024
Revised: 28 June 2024
Accepted: 19 August 2024
First Online: 4 September 2024
Compliance with ethical standards
:
: The scientific guarantor of this publication is R.G.
: The first authors of this manuscript declare relationships with the following companies: J.S.K. is a cofounder and shareholder of Bonescreen GmbH. See . 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.
: One of the authors has significant statistical expertise.
: Prospective data (NAKO/SHIP): Written informed consent was obtained from all subjects (patients) in this study. Internal data: Written informed consent was not required for this study because we used existing MRI images from the Clinic “Rechts der Isar.” We anonymized the MRI. According to German law (Bavarian hospital law), we can do research with existing images without written consent internally if we follow the guidelines of patient privacy and ethics. We adhere to the German ethics standards for medical research on patient data generated by normal clinical routine.
: The ethics committee of the Technical University Munich approved our German-law-compliant studies (593/21 S-NP). Study-specific ethics approval is not required as long we stay in the rules of the Bavarian hospital law for retrospective data analysis and privacy protection.
: Some study subjects or cohorts have been previously reported. NAKO and SHIP are shared large datasets and have multiple studies. The internal data was used in studies by Schläger et al in: ; . The internal data consists of two data collections. One of the data collections did not collect the age, sex, height, and weight information from the scanner and cannot be retrieved. The transformation from T1-weighted Dixon + T2w FSE to T1w and STIR imaging is a novel contribution. All experiments and deep learning models presented in the study are newly developed.
: