Table 4.
Methodological quality assessment using the RQS
| Criterion | Points range | Median points (percentage) |
|---|---|---|
| Image protocol quality | 0–2 | 1 (50%) |
| Multiple segmentations | 0–1 | 1 (50%) |
| Phantom study on all scanners | 0–1 | 0 (0%) |
| Imaging at multiple time points | 0–1 | 0 (0%) |
| Feature reduction or adjustment for multiple testing | −3 to 3 | 3 (100%) |
| Multivariable analysis with non-radiomics features | 0–1 | 0 (0%) |
| Detect and discuss biological correlates | 0–1 | 0 (0%) |
| Cut-off analyses | 0–1 | 0 (0%) |
| Discrimination statistics | 0–2 | 1 (50%) |
| Calibration statistics | 0–2 | 0 (0%) |
| Prospective study registered in a trial database | 0–7 | 0 (0%) |
| Validation | −5 to 5 | 2 (40%) |
| Comparison to ‘gold standard' | 0–2 | 0 (0%) |
| Potential clinical utility | 0–2 | 0 (0%) |
| Cost-effectiveness analysis | 0–1 | 0 (0%) |
| Open science and data | 0–4 | 0 (0%) |
| Total | −8 to 36 | 11 (30.6%) |
Median points percentage was calculated by dividing the scored points by the ideal points