Background: The potential of Segment Anything Model 2 (SAM2) for 3D medical image segmentation via video-stream processing is currently constrained by its reliance on manual prompts. While existing research employs auxiliary mode... ...

请注册登录后继续浏览