Skip to main content

Table 2 Basic adherence rate according to the six key domains

From: A systematic review reporting quality of radiomics research in neuro-oncology: toward clinical utility and quality improvement using high-dimensional imaging features

 Basic adherence rate
Total 16 items37.1%
Domain 1: Protocol quality and stability in image and segmentation32.3%
 Protocol quality51 (100%)
 Test-retest1 (2%)
 Phantom study0 (0%)
 Multiple segmentation14 (27.4%)
Domain 2: Feature selection and validation81.4%
 Feature reduction or adjustment of multiple testing48 (94.1%)
 Validation35 (68.6%)
Domain 3: Biologic/clinical validation and utility39.2%
 Multivariate analysis with non-radiomics features32 (62.7%)
 Biologic correlates28 (74.5%)
 Comparison to ‘gold standard’19 (37.2%)
 Potential clinical utility1 (2%)
Domain 4: Model performance index45.1%
 Discrimination statistics49 (96.1%)
 Calibration statistics7 (13.7%)
 Cut-off analysis13 (25.5%)
Domain 5: High level of evidence2%
 Prospective study2 (3.9%)
 Cost-effective analysis0 (0%)
Domain 6: Open science and data3 (5.9%)