The authors of this study aimed to systematically review radiomic feature reproducibility and predictive model validation strategies in studies that deal with CT and MRI radiomics of bone and soft-tissue sarcomas. The review consisted of 278 papers, forty-nine of which were published between 2008 and 2020. The authors found that the issues of radiomic feature reproducibility and model validation varied largely among the studies dealing with musculoskeletal sarcomas, determining that they should be addressed in future investigations in order to bring the field of radiomics to the clinical stage.
Key points
- Radiomic studies focused on CT and MRI of musculoskeletal sarcomas were reviewed.
- Feature reproducibility analysis and model validation strategies varied largely among these studies.
- Radiomic feature reproducibility was assessed in less than half of the studies.
- Only 10% of the studies included an independent clinical validation of the model.
Authors: Salvatore Gitto, Renato Cuocolo, Domenico Albano, Francesco Morelli, Lorenzo Carlo Pescatori, Carmelo Messina, Massimo Imbriaco & Luca Maria Sconfienza
