Deep Learning Model Successfully Segments Robotic Pancreaticojejunostomy Videos

Researchers developed a deep learning model for automated segmentation of robotic pancreaticojejunostomy videos. The model achieved an accuracy of 88.01% and mean per-class f1 scores of 85.34%, proving to be effective in identifying the tasks involved in the procedure. Future work will focus on skills assessment and outcome prediction based on the segmentation results.