Fig. 5: Multiparmetric analyses and all-around US score using a support vector regression (SVR) approach. | Nature Communications

Fig. 5: Multiparmetric analyses and all-around US score using a support vector regression (SVR) approach.

From: 3D multiparametric ultrasound imaging of steatotic liver disease in a study with male rats

Fig. 5

a Scatter plots of showing correlations between the fat percentage and 3D USI indices (n = 37 for each index): UFD indices (weight-normalized vessel volume occupancy (wnVVO), weight-normalized fractional moving blood volume (wnFMBV), vessel bifurcation density (VBD), and sum of angles metric (SOAM)), ATI index (estimated attenuation coefficient (EAC)), and ASQ index (focal disturbance ratio (FDR)). Two-sided Pearson’s and Spearman’s correlation tests were performed (***p < 0.001). b All-around US score intergrating five 3D USI indices (wnVVO, wnFMBV, VBD, SOAM, and EAC) using SVR (n = 37). Two-sided Pearson’s and Spearman’s correlation tests were performed (***p < 0.001). In the classifications of grade 0 vs. 1 and grade 1 vs. 2, estimating 95% confidence intervals is excluded due to insufficient data. PC Pearson’s coefficient, SC Spearman’s coefficient, ROC receiver operating characteristic, TPR true positive rate, FPR false positive rate, Se sensitivity, Sp specificity, and BAC balanced accuracy. The results for each index were derived only once from each subject (i.e., biological replicate). Exact p-values are provided in source data. Source data are provided as a Source Data file.

Back to article page