Fig. 2
From: Automated interpretation of cardiotocography using deep learning in a nationwide multicenter study
Model performance. The figure shows the performance metrics of the model evaluated on internal and external test datasets using AUC and PRC curves. The left plot represents the AUC (Area Under the Receiver Operating Characteristic curve), indicating the model’s discrimination ability across internal and external test sets. The right plot illustrates the PRC (Precision-Recall Curve), highlighting the model’s performance in detecting abnormal cases. The accompanying table summarizes key metrics, including AUC, PRC, specificity (SPE), and negative predictive value (NPV), for each dataset. Consistent performance across internal and external tests demonstrates the model’s generalizability.
