Figure 1: Overview of Surgment, a web-based system that helps expert surgeons create visual questions and feedback based on surgery videos to enhance video-based surgery learning. 1○ Surgment is powered by a surgery scene segmentation pipeline (SegGPT+SAM), which generates an accurate understanding of the surgery scene composition. Based on the scene segmentation result, Surgment has two key design features, namely 2 ○ A search-by-mask tool, which enables surgeons to quickly identify image frames by adjusting the position, size, and shape of the masks. 3 ○ A quiz-maker tool, which enables surgeons to create visual questions and feedback that target specific anatomical structures and surgical tools.