Table 2. Optimal Cutoff Values for Differentiation between the lesion and the normal rectal wall.
| Parameters | Cutoff value | AUC (95%CI) | Sensitivity | specificity | P value |
|---|---|---|---|---|---|
| Standard-ADC(×10−3 mm2/s) | 0.976 | 1.000 (0.969–1.000) | 100% | 100% | 1.000a |
| Slow ADC-Mono(×10−3 mm2/s) | 0.773 | 1.000 (0.969–1.000) | 100% | 100% | 0.037b |
| Fast ADC-Mono (×10−3 mm2/s) | 18.267 | 0.520 (0.426–0.613) | 96.61% | 23.73% | 0.226c |
| Fraction-Mono | 0.348 | 0.911 (0.845–0.956) | 96.61 | 77.97 | 0.037d |
| Slow ADC-bi (×10−3 mm2/s) | 0.729 | 0.975 (0.928–0.995) | 91.53% | 94.92 | 0.226e |
| Fast ADC-bi(×10−3 mm2/s) | 60.733 | 0.786 (0.700–0.856) | 81.36% | 66.10% | 0.827 f |
| Fraction-bi | 0.32 | 0.558 (0.464–0.649) | 47.46 | 71.19 | |
| α value | 0.806 | 0.751 (0.663–0.826) | 100% | 49.15 | |
| DDC (×10−3 mm2/s) | 1.124 | 0.980 (0.935–0.997) | 98.31% | 94.92% |
AUC (area under the curve), CI (confidence interval), superscript letters of a(Standard ADC vs Slow ADC-Mono), b(Standard ADC vs Slow ADC-bi), c(Standard ADC vs DDC), d(Slow ADC-Mono vs Slow ADC-bi), e(Slow ADC-Mono vs DDC), f(Slow ADC –bi vs DDC). P values < (0.05/6) were considered significant after Bonferroni correction.