Table 2. Association between PO-IoU and KL grade. Unless noted, statistics are computed on the individual radiographs (N = 8,252).
| Analysis | Statistic | p |
| Linear regression | R2 = 0.048 | 1.8 × 10−90 |
| Spearman | ρ = −0.221 | 5.2 × 10−92 |
| Kendall | τb = −0.168 | 1.2 × 10−92 |
| Kruskal–Wallis | H = 635.4, ε2 = 0.077 | 3.4 × 10−136 |
| KL 0–1 vs. KL 3–4 | AUC = 0.713 | 3.2 × 10−125 |
| Grade-mean regression | R2 = 0.819 | 3.5 × 10−2 |
Linear-regression slope = −0.031; Kruskal–Wallis df = 4; AUC 95% CI [0.699, 0.727]. The grade-mean regression is fitted to five aggregated points. Pairwise post-hoc tests (Mann–Whitney, Bonferroni-corrected over ten comparisons) are significant for nine of the ten grade pairs, the sole exception being KL 0 versus KL 1; KL 3 and KL 4 differ significantly (p = 1.4 × 10−19, Cliff’s δ = 0.379) with KL 4 lowest. Grade-wise bootstrap 95% confidence intervals of the mean (5,000 resamples) are disjoint from KL 2 onward.