Table 5.
ROC curves of Kep, TTP and ADC in differential diagnosis.
| ROC analysis | ||||||
|---|---|---|---|---|---|---|
| Identifying group | Parameters | Cutoff | AUC (95%CI) | Accuracy | Sensitivity | Specificity |
| Luminal A vs. TNBC | Kep | 0.892 | 0.675 (0.540–0.809) | 0.656 | 0.559 | 0.767 |
| TTP | 0.582 | 0.755 (0.627–0.884) | 0.766 | 0.765 | 0.767 | |
| ADC | 0.719 | 0.667 (0.534–0.800) | 0.625 | 0.353 | 0.933 | |
| Combined | 0.527 | 0.811 (0.708–0.915) | 0.750 | 0.735 | 0.767 | |
| Luminal B vs. HER-2-enriched | ADC | 0.778 | 0.702 (0.592–0.812) | 0.554 | 0.889 | 0.432 |
| Luminal B vs. TNBC | TTP | 0.583 | 0.699 (0.596–0.802) | 0.676 | 0.765 | 0.635 |
| HER-2-enriched vs. TNBC | TTP | 0.582 | 0.647 (0.501–0.794) | 0.705 | 0.765 | 0.630 |
| ADC | 0.759 | 0.763 (0.644–0.882) | 0.705 | 0.500 | 0.963 | |
| Combined | 0.405 | 0.793 (0.681–0.904) | 0.754 | 0.882 | 0.593 | |
Combined indicate the predicted model based on the combination of the parameters. ROC, Receiver operating characteristic; AUC, Area under the curve; CI, Confidence interval; TNBC, Triple-negative breast cancer; ADC, Apparent diffusion coefficient.