CLEAR item#54

“Sharing radiomic feature data. Share selected radiomic feature data along with clinical variables or labels with the public, if possible (i.e., in accordance with the regulatory constraints of the institute). Specify the reason if the radiomic feature data are not available.” [1] (from the article by Kocak et al.; licensed under CC BY 4.0)

Reporting examples for CLEAR item#54

Example#1. “The data and the R code that support the findings of this study are available at https://github.com/harveerar/SciRepEndometrial2020.” [2] (from the article by Veeraraghavan et al.; licensed under CC BY 4.0)

Example#2. “All the data described here as the “UPenn-GBM” collection, are available from the publicly available repository of The Cancer Imaging Archive (TCIA) at: https://doi.org/10.7937/TCIA.709X-DN49. Data availability per subject can also be found and downloaded from the TCIA repository.” [3] (from the article by Bakas et al.; licensed under CC BY 4.0)

Example#3. “[…] can be downloaded from https://github.com/ehodneland/RadioGenomicsEC. The same repository also contains code for the training of the network, as well as for extraction and clustering of radiomic features.” [4] (from the article by Hoivik et al.; licensed under CC BY 4.0)

Explanation and elaboration of CLEAR item#54

Item#54 emphasizes the importance of open sharing of radiomics data. Any radiomics study should make the extracted features and any clinical variables available to the scientific community. Data sharing can help to generate larger datasets, thereby increasing the statistical power of analyses [5]. In addition, datasets of features obtained using the same extraction method can be merged to form larger datasets, ensuring anonymization of patients. The creation of radiomics biobanks would also be desirable in terms of linking with genetics or tissue biobanks [6]. However, a 2015 ESR survey showed that in 80% of the cases, access to these biobanks is restricted [7]. This may be due to compliance with regulatory restrictions, which vary from country to country. Transparency in the sharing of these data can facilitate collaboration, validation of results and the development of more robust predictive models.

References

  1. Kocak B, Baessler B, Bakas S, et al (2023) CheckList for EvaluAtion of Radiomics research (CLEAR): a step-by-step reporting guideline for authors and reviewers endorsed by ESR and EuSoMII. Insights Imaging 14:75. https://doi.org/10.1186/s13244-023-01415-8
  2. Veeraraghavan H, Friedman CF, DeLair DF, et al (2020) Machine learning-based prediction of microsatellite instability and high tumor mutation burden from contrast-enhanced computed tomography in endometrial cancers. Sci Rep 10:17769. https://doi.org/10.1038/s41598-020-72475-9
  3. Bakas S, Sako C, Akbari H, et al (2022) The University of Pennsylvania glioblastoma (UPenn-GBM) cohort: advanced MRI, clinical, genomics, & radiomics. Sci Data 9:453. https://doi.org/10.1038/s41597-022-01560-7
  4. Hoivik EA, Hodneland E, Dybvik JA, et al (2021) A radiogenomics application for prognostic profiling of endometrial cancer. Commun Biol 4:1–12. https://doi.org/10.1038/s42003-021-02894-5
  5. Purushotham S, Meng C, Che Z, Liu Y (2018) Benchmarking deep learning models on large healthcare datasets. J Biomed Inform 83:112–134. https://doi.org/10.1016/j.jbi.2018.04.007
  6. Coppola L, Cianflone A, Grimaldi AM, et al (2019) Biobanking in health care: evolution and future directions. J Transl Med 17:172. https://doi.org/10.1186/s12967-019-1922-3
  7. European Society of Radiology (ESR) (2015) ESR Position Paper on Imaging Biobanks. Insights Imaging 6:403–410. https://doi.org/10.1007/s13244-015-0409-x

Back Next