Table 3.
Summary of classification task performances.
| Dataset | Fibrosis 0–1 vs. 2–4 |
Fibrosis 0–2 vs. 3–4 |
||
|---|---|---|---|---|
| Median AUC (95% CI) | Minimal features (n) | Median AUC (95% CI) | Minimal features (n) | |
| LM | 0.73 (0.058) | 13 | 0.77 (0.053) | 12 |
| LMC | 0.77 (0.048) | 9 | 0.78 (0.044) | 10 |
| C | 0.75 (0.056) | 7 | 0.78 (0.051) | 7 |
| Improvement by adding (LM) to (C) | 0.022 | 0.004 | ||
| Performance of (LM alone) as % of (C) | 98.1 | 98.3 | ||
C, clinical variables used as predictors; LM, lipids and metabolites used as predictors; LMC, lipids, metabolites, and clinical variables used as predictors.