Skip to main content

Table 6 Predictive performance of combined model, radiomics model, and clinicoradiological model

From: Machine learning-based CT radiomics approach for predicting WHO/ISUP nuclear grade of clear cell renal cell carcinoma: an exploratory and comparative study

Model Combined model Radiomics model Clinicoradiological model
Training cohort Validation cohort Testing cohort Training cohort Validation cohort Testing cohort Training cohort Validation cohort Testing cohort
AUC 0.887 [0.798–0.952] 0.859 [0.748–0.935] 0.828 [0.731–0.929] 0.860 [0.759–0.963] 0.824 [0.736–0.915] 0.787 [0.710–0.892] 0.752 [0.649–0.870] 0.703 [0.592–0.844] 0.637 [0.511–0.769]
Accuracy (%) 85.24 [76.75–90.14] 82.76 [75.99–91.36] 81.62 [74.18–92.45] 83.61 [75.85–92.65] 77.14 [69.38–86.05] 73.42 [61.63–82.66] 69.33 [56.54–83.76] 62.87 [51.74–75.36] 56.35 [42.48–69.49]
Sensitivity (%) 89.77 [80.63–96.98] 84.94 [76.72–92.49] 85.56 [77.82–91.58] 84.78 [74.36–93.74] 78.72 [64.35–91.48] 82.15 [73.74–91.92] 71.86 [59.43–83.82] 73.89 [65.95–83.47] 65.78 [46.89–80.87]
Specificity (%) 84.47 [72.60–91.65] 83.42 [74.32–92.18] 78.01 [69.45–85.74] 82.67 [74.35–93.14] 80.33 [69.74–92.26] 76.48 [63.36–89.17] 70.42 [61.36–83.47] 66.86 [56.63–78.10] 64.23 [52.36–80.62]
PPV (%) 82.34 [73.87–91.59] 81.52 [73.67–87.59] 78.36 [70.16–86.47] 77.79 [69.16–85.97] 75.98 [68.30–83.57] 74.82 [65.87–83.64] 68.24 [56.25–81.39] 65.71 [52.63–78.78] 60.99 [48.34–73.76]
NPV (%) 87.65 [79.48–97.86] 86.77 [78.43–91.17] 85.60 [79.98–90.35] 83.90 [72.74–95.46] 80.43 [71.32–89.91] 82.62 [71.64–92.35] 74.22 [65.61–86.34] 70.98 [61.05–82.77] 73.17 [60.05–87.47]
  1. Data in parentheses are 95% CIs