Fig. 3: Segmentation accuracy and diagnostic performance across models. | Nature Communications

Fig. 3: Segmentation accuracy and diagnostic performance across models.

From: Automated MRI system for clinically significant prostate cancer detection development validation and real-world implementation

Fig. 3

A–C Prostate-gland Dice across nnUNet, nnSAM and LightM-UNet. D Schematic of gland segmentation. E Lesion-level Dice across models. F ROC curves for lesion classification. G–J Patient-level ROC curves for PI-RADS and ProAI across datasets; AUCs were compared using DeLong’s test. All statistical tests were two-sided with P < 0.05 considered significant unless stated. ROC receiver operating characteristic, AUC area under the ROC curve, TCIA The Cancer Imaging Archive. Source data are provided as a Source Data file.

Back to article page