Fig. 7: Predicted segmentations and ground truth of four patients. | Communications Medicine

Fig. 7: Predicted segmentations and ground truth of four patients.

From: Patient-specific prostate segmentation in kilovoltage images for radiation therapy intrafraction monitoring via deep learning

Fig. 7

Comparison of the conditional generative adversarial network segmentation (dashed yellow) and ground truth (solid red) at imager angles of 90°, 120°, 150°, and 180°. The best and worst performing patient in each dataset, based on the mean Dice similarity coefficient (DSC), are shown. The DSC is reported for each image. AP/LAT, anterior–posterior/lateral; SI, superior–inferior.

Back to article page