Table 2

Performance of predict models for predicting efficacy of combined immunotherapy in patients with advanced breast cancer

ValueTraining setTesting set
Clinical modelRadiomics modelCombined modelClinical modelRadiomics modelCombined model
AUC (95% CI)0.672 (0.588 to 0.756)0.994 (0.988 to 1.000)0.997 (0.993 to 1.000)0.634 (0.495 to 0.772)0.920 (0.824 to 1.000)0.961 (0.885 to 1.000)
SEN0.5570.9510.9840.5000.8750.958
SPE0.7180.9820.9730.8001.0001.000
ACC0.6610.9710.9770.6960.9570.986
PPV0.5230.9670.9520.5711.0001.000
NPV0.7450.9730.9910.7500.9380.978
  • ACC, accuracy; AUC, area under the receiver operating curve; NPV, negative predictive value; PPV, positive predictive value; SEN, sensitivity; SPE, specificity.