Search

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. [ABSTRACT FROM AUTHOR]

Objectives: Whole-body MR imaging is increasingly utilised; although for lung dedicated sequences are often not included, the chest is typically imaged. Our objective was to determine the clinical utility of lung volumes derived from non-dedicated MRI sequences in the population-based KORA-FF4 cohort study.Methods: 400 subjects (56.4 ± 9.2 years, 57.6% males) underwent whole-body MRI including a coronal T1-DIXON-VIBE sequence in inspiration breath-hold, originally acquired for fat quantification. Based on MRI, lung volumes were derived using an automated framework and related to common predictors, pulmonary function tests (PFT; spirometry and pulmonary gas exchange, n = 214) and obstructive lung disease.Results: MRI-based lung volume was 4.0 ± 1.1 L, which was 64.8 ± 14.9% of predicted total lung capacity (TLC) and 124.4 ± 27.9% of functional residual capacity. In multivariate analysis, it was positively associated with age, male, current smoking and height. Among PFT indices, MRI-based lung volume correlated best with TLC, alveolar volume and residual volume (RV; r = 0.57 each), while it was negatively correlated to FEV1/FVC (r = 0.36) and transfer factor for carbon monoxide (r = 0.16). Combining the strongest PFT parameters, RV and FEV1/FVC remained independently and incrementally associated with MRI-based lung volume (β = 0.50, p = 0.04 and β = - 0.02, p = 0.02, respectively) explaining 32% of the variability. For the identification of subjects with obstructive lung disease, height-indexed MRI-based lung volume yielded an AUC of 0.673-0.654.Conclusion: Lung volume derived from non-dedicated whole-body MRI is independently associated with RV and FEV1/FVC. Furthermore, its moderate accuracy for obstructive lung disease indicates that it may be a promising tool to assess pulmonary health in whole-body imaging when PFT is not available.Key Points: • Although whole-body MRI often does not include dedicated lung sequences, lung volume can be automatically derived using dedicated segmentation algorithms • Lung volume derived from whole-body MRI correlates with typical predictors and risk factors of respiratory function including smoking and represents about 65% of total lung capacity and 125% of the functional residual capacity • Lung volume derived from whole-body MRI is independently associated with residual volume and the ratio of forced expiratory volume in 1 s to forced vital capacity and may allow detection of obstructive lung disease. [ABSTRACT FROM AUTHOR]