HomeNewsEvaluation of two deep learning automatic segmentation methods for estimating retroperitoneal sarcoma volume on CT scans

Evaluation of two deep learning automatic segmentation methods for estimating retroperitoneal sarcoma volume on CT scans

A recent study from an Italian group retrospectively assessed the performance of two fully automated deep learning networks (ENet and ERFNet) for retroperitoneal sarcoma (RPS) segmentation and volume analysis, comparing them with the manual approach. Based on abdominal CT examinations, both deep learning automatic segmentation methods provided a faster assessment of RPS volume over manual segmentation, with ENet performing better than ERFNet and ERFNet proving less time-consuming than ENet.