Skip to main content

Table 3 The performance of three prediction models in the training and validation cohorts

From: Feasibility of a clinical-radiomics combined model to predict the occurrence of stroke-associated pneumonia

Model

Cohort

AUC (95%CI)

ACC

Sensitivity

Specificity

NPV

PPV

Radiomics Prediction model

Training Cohort

0.660(0.546–0.766)

0.779

0.417

0.826

0.916

0.238

Validation Cohort

0.646(0.541–0.756)

0.700

0.533

0.733

0.887

0.286

Clinical Prediction model

Training Cohort

0.785(0.673–0.889)

0.759

0.667

0.768

0.939

0.300

Validation Cohort

0.736(0.629–0.837)

0.811

0.417

0.872

0.907

0.333

Combined Prediction model

Training Cohort

0.859(0.759–0.936)

0.817

0.750

0.826

0.962

0.360

Validation Cohort

0.830(0.758–0.896)

0.789

0.800

0.787

0.952

0.429