A practical guide to the implementation of AI in orthopaedic research-Part 5: Data management
Review article, 2025
While the magnitude and types of data available to orthopaedic researchers are steadily growing, standardized and efficient data management workflows for orthopaedic research using artificial intelligence (AI) are currently lacking. This work introduces essential principles and best practices for planning, collecting, storing, processing, labelling and governing data in AI-based orthopaedic research. The various domains of available data quality guidelines for medical AI research are reviewed and discussed in terms of their adaptability to orthopaedic research datasets. In addition, future areas of improvement, such as registry development, the potential of synthetic data and gradual transition to continuous data streams for AI applications, are outlined.Level of Evidence Level V.
methods
machine learning
causal inference
data analysis
artificial intelligence
Author
Balint Zsidai
University of Gothenburg
Felix Oettl
University of Zürich
James A. Pruneski
Tripler Army Medical Center
Gergely Panics
FIFA Medical Centre of Excellence (FMCE)
Semmelweis University
Philipp W. Winkler
University of Gothenburg
Eric Hamrin Senorski
University of Gothenburg
Michael T. Hirschmann
Canton Hospital Basel-Land
Yinan Yu
Chalmers, Computer Science and Engineering (Chalmers), Functional Programming
Robert Feldt
Chalmers, Computer Science and Engineering (Chalmers), Software Engineering (Chalmers)
Kristian Samuelsson
University of Gothenburg
Journal of Experimental Orthopaedics
2197-1153 (eISSN)
Vol. 12 4 e70581Subject Categories (SSIF 2025)
Orthopaedics
Artificial Intelligence
DOI
10.1002/jeo2.70581
PubMed
41416245