A robot trained solely through imitation learning by watching expert surgeons on video has successfully performed complex surgeries with human-like precision.
This milestone brings robotic surgery closer to full autonomy, where robots could perform procedures without human guidance.
Researchers at Johns Hopkins University emphasize that imitation learning eliminates the need for programming individual surgical maneuvers.
Axel Krieger, a senior researcher, calls it “magical” that feeding the model camera input enables it to predict robotic movements for surgery.
Stanford University researchers trained a da Vinci Surgical System robot in tasks like needle handling, tissue lifting, and suturing.
The robot performed these tasks with skill levels comparable to those of experienced human surgeons.
The training model combines ChatGPT-like machine learning architecture with robotic kinematics, converting motion angles into mathematical expressions.