Incorporating radiomics into clinical trials: expert consensus endorsed by the European Society of Radiology on considerations for data-driven compared to biologically driven quantitative biomarkers

Abstract

Existing quantitative imaging biomarkers (QIBs) are associated with known biological tissue characteristics and follow a well-understood path of technical, biological and clinical validation before incorporation into clinical trials. In radiomics, novel data-driven processes extract numerous visually imperceptible statistical features from the imaging data with no a priori assumptions on their correlation with biological processes. The selection of relevant features (radiomic signature) and incorporation into clinical trials therefore requires additional considerations to ensure meaningful imaging endpoints. Also, the number of radiomic features tested means that power calculations would result in sample sizes impossible to achieve within clinical trials. This article examines how the process of standardising and validating data-driven imaging biomarkers differs from those based on biological associations. Radiomic signatures are best developed initially on datasets that represent diversity of acquisition protocols as well as diversity of disease and of normal findings, rather than within clinical trials with standardised and optimised protocols as this would risk the selection of radiomic features being linked to the imaging process rather than the pathology. Normalisation through discretisation and feature harmonisation are essential pre-processing steps. Biological correlation may be performed after the technical and clinical validity of a radiomic signature is established, but is not mandatory. Feature selection may be part of discovery within a radiomics-specific trial or represent exploratory endpoints within an established trial; a previously validated radiomic signature may even be used as a primary/secondary endpoint, particularly if associations are demonstrated with specific biological processes and pathways being targeted within clinical trials. Key Points • Data-driven processes like radiomics risk false discoveries due to high-dimensionality of the dataset compared to sample size, making adequate diversity of the data, cross-validation and external validation essential to mitigate the risks of spurious associations and overfitting. • Use of radiomic signatures within clinical trials requires multistep standardisation of image acquisition, image analysis and data mining processes. • Biological correlation may be established after clinical validation but is not mandatory.

Radiomic Score as a Potential Imaging Biomarker for Predicting Survival in Patients With Cervical Cancer

“…Two-dimensional features may increase the robustness of features compared with three-dimensional features. Seventh, the biological interpretation of radiomics remains an open question warranting further investigation (50). Finally, this is a single-institution study that needs external validation of the findings.…”

The Way of the Future: Personalizing Treatment Plans Through Technology

“…The growing field of quantitative imaging biomarkers recognizes the need for standards that address each of the stages of imaging technology, including the imaging device itself ("scanner"), the image reconstruction process, and the method of image quantification. [2][3][4] Each of these elements contributes to the overall performance of the imaging biomarker and has to be considered in the optimization process. In breast cancer care, much is being learned in the neoadjuvant treatment setting where the status of the tumor can be monitored during systemic treatment.…”

Bone Metastases Are Measurable: The Role of Whole-Body MRI and Positron Emission Tomography

“…The Imaging Biomarker Standardization Initiative has harmonized performance of radiomics software packages to allow for its robust and reproducible use (157). Recently, consensus recommendations for considerations on the use of radiomics (both PET, CT, and MRI) in clinical trials have been proposed (158). Deep learning techniques, which do not require extraction of predefined features seem particularly promising for segmentation purposes of PET-avid bone metastases (159,160).…”