U-Net-based Pancreas Tumor Segmentation from Abdominal CT Images

There is no doubt that pancreatic cancer one of the most deadly types cancer. In order to diagnose and stage tumors, computed tomography (CT) widely used. However, manual segmentation volumetric CT scans a time-consuming subjective process. It has been shown U-Net model highly effective for semantic segmentation, although several deep learning models have proposed. this study, we propose U-Net-based method tumor from abdominal images demonstrate its simplicity effectiveness. Using architecture, pancreas segmented slices in first stage, while tumors are masked second stage. For validation set NIH dataset according proposed method's dice scores, performance was outstanding, demonstrating potential identify efficiently accurately.