Fig. 5: Performance evaluation of the PAN-VIQ model on a prospective dataset of 202 cases in 2024. | npj Digital Medicine

Fig. 5: Performance evaluation of the PAN-VIQ model on a prospective dataset of 202 cases in 2024.

From: A clinically validated 3D deep learning approach for quantifying vascular invasion in pancreatic cancer

Fig. 5

A Confusion matrices showing the detailed classification results of the PAN-VIQ model for five major vessels: CA, CHA, SMA, SMV, and PV. The heatmap colors represent the number of samples, ranging from red (higher sample counts) to blue (lower sample counts).Performance comparison among the PAN-VIQ model, senior radiologists, and junior radiologists across different evaluation metrics, including accuracy, precision, recall, F1-score, and specificity. B The radar charts illustrate the overall performance distribution. C Bar graphs provide a detailed metric-by-metric comparison. The results indicate that the PAN-VIQ model outperforms junior radiologists in all metrics and achieves performance comparable to senior radiologists in certain aspects. Abbreviations: CA celiac artery, CHA common hepatic artery, SMA superior mesenteric artery, SMV superior mesenteric vein, PV portal vein.

Back to article page