Johns Hopkins University researchers found that a robot trained by imitation learning using AI could perform three fundamental surgical tasks, manipulating the needle, lifting body tissue, and suturing, skillfully as human doctors.
The research team, which included Stanford University researchers, used a model that combined imitation learning and AI to train the da Vinci Surgical System robot. The model was fed hundreds of videos from wrist cameras placed on the da Vinci robots during surgeries. The key was training the model to perform relative movements rather than absolute actions.
Axel Krieger, Assistant Professor, Department of Mechanical Engineering, said, “We believe this marks a significant step forward toward a new frontier in medical robotics.” Before this breakthrough, it could take many years to program a robot to perform even a simple aspect of a surgical procedure.