Table 1.
Prediction accuracy of linear unmixing and difference analysis in detecting each grade of steatosis.
| Cross‐sectional steatosis grade classification (n = 43) | Grades 1–3 (n = 29) versus grade 0 (n = 14) | Grades 2–3 (n = 18) versus grade 0–1 (n = 25) | Grade 3 (n = 9) versus grades 0–2 (n = 34) | |
|---|---|---|---|---|
| Linear unmixing | cutoff | 17.3 | 42.9 | 252.9 |
| AUROC (95%CI) | 0.90 (0.81–0.98) | 0.99 (0.96–1.01) | 0.98 (0.94–1.02) | |
| Se | 72% | 100% | 89% | |
| Sp | 100% | 92% | 97% | |
| PPV | 100% | 90% | 89% | |
| NPV | 64% | 100% | 97% | |
| Difference analysis | cutoff | 0.74 | 0.97 | 1.01 |
| AUROC (95% CI) | 0.93 (0.86–1.02) | 0.80 (0.67–0.94) | 0.83 (0.68–0.98) | |
| Se | 100% | 67% | 89% | |
| Sp | 79% | 84% | 82% | |
| PPV | 91% | 75% | 57% | |
| NPV | 100% | 78% | 97% | |
| Difference analysis versus linear unmixing P value (sample size not enough) | 0.48 | 0.008 | 0.06 |
Se, sensitivity; Sp, specificity; NPV, negative predictive value; and PPV, positive predictive value.