Skip to main content

Table 2 Accuracy and predictive value between six models

From: A clinical-radiomics nomogram for the preoperative prediction of lymph node metastasis in colorectal cancer

Training cohortAUC95% CISensitivitySpecificityAccuracyPPVNPV
Clinical features0.7127[0.6876–0.7367]0.54660.78440.67850.67070.6829
Lesion radiomics0.6007[0.5756–0.6273]0.60420.55880.57900.52280.6383
Lymph node radiomics0.6441[0.6171–0.6681]0.43630.78530.63020.61060.6564
Clinical-lesion radiomics0.7299[0.7063–0.7524]0.59310.79510.70530.69840.7095
Clinical-lymph node radiomics0.7519[0.7288–0.7738]0.56370.82550.70920.62820.7500
Clinical-lesion-lymph node radiomics0.7606[0.7373–0.7833]0.64950.79020.72770.71240.7381
Validation cohortAUC95% CISensitivitySpecificityAccuracyPPVNPV
Clinical features0.7075[0.6448–0.7650]0.69120.63370.65910.59870.7219
Lesion radiomics0.5276[0.4634–0.5927]0.30880.77910.57140.52500.5877
Lymph node radiomics0.6500[0.5829–0.7133]0.50740.74420.63960.61060.6564
Clinical-lesion radiomics0.7055[0.6459–0.7637]0.69120.63950.66230.60260.7237
Clinical-lymph node radiomics0.7359[0.6750–0.7899]0.72060.66280.68830.62820.7500
Clinical-lesion-lymph node radiomics0.7509[0.6901–0.8071]0.60290.84300.73700.75230.7286
  1. AUC area under the curve, PPV positive predictive value, NPV negative predictive value