Table 1.
Patient characteristics.
Fig 1.
Box plots of the CPAa (A) and SWVb (B).
acollagen proportionate area; bshear wave velocity.
Fig 2.
ROCa curves dichotomizing liver fibrosis stages.
To dichotomize F1 versus F2–4 (A), the AUROCb for the CPAc was 0.9349 (95% confidence interval: 0.8943–0.9755) and SWVd was 0.8434 (0.7762–0.9105) (CPA versus SWV, P = 0.0063). For F1, 2 versus F3, 4 (B), the CPA was 0.9436 (0.9091–0.9781); SWV was 0.8997 (0.8444–0.9551) (P = 0.1587). For F1–3 versus F4 (C), the CPA was 0.8647 (0.7944–0.9349); SWV was 0.9036 (0.8499–0.9573) (P = 0.2585). APRIe was incorporated too. areceiver operating characteristic; barea under ROC curves; ccollagen proportionate area; dshear wave velocity; easpartate transaminase-to-platelet ratio index.
Table 2.
Liver fibrosis dichotomization using the CPA and SWV.
Table 3.
Multiple regression analyses for the CPAa.
Table 4.
Multiple regression analyses for the SWVd.
Fig 3.
Scatter plot between the CPAa (Y-axis) and SWVb (X-axis).
For CPA prediction by using SWV through univariate linear regression, the CPA (%) = -4.768 + SWV (m/s) × 10.184, as a line of best fit (R2 = 0.455, P < 0.001). Eight of 137 cases (5.8%) exceeded the 95% confidence intervals. acollagen proportionate area; bshear wave velocity.