Researchers from John Hopkins University have successfully trained a robot to perform complex surgical tasks at the same level as human surgeons, simply by watching seasoned doctors perform surgeries. At the recently held Conference on Robot Learning in Munich, a popular event for robotics and machine learning, researchers said that they used imitation learning to train the da Vinci Surgical System robot to perform "three fundamental tasks" that include manipulating a needle, lifting body tissue and suturing. They also said that the technique eliminates the need to program robots with individual moves and brings robotic surgery a step closer to autonomy. "It's really magical to have this model and all we do is feed it camera input and it can predict the robotic movements needed for surgery," said senior author Axel Krieger, an assistant professor in JHU's Department of Mechanical Engineering. He added that "the model is so good learning things we haven't taught it. Like if it drops the needle, it will automatically pick it up and continue. This isn't something I taught it do." While robotic or robot-assisted surgery isn't new, they are often guided by doctors using joystick-like controllers. However, robots are now capable of understanding these moves and even fixing mistakes by themselves without human intervention. The team working on these robots said that the approach to train the surgical robot is similar to how large language models like ChatGPT are developed. However, instead of words and images, the model was trained using videos recorded from wrist cameras placed on the arms of the da Vinci robots during surgical procedures. Before this advancement, programming a surgical robot took a lot of time as it required hand-coding every single step. According to Jio Woong Brian Kim, a postdoctoral researcher working with the team, the system has been developed in such a way that doctors can talk to the robot just like you would do with a real surgical resident and even say things like 'Move left', 'Move right' and 'Do this task'. While this may sound impressive, experts say that the stakes are pretty high since a small error may lead to death, so we may have to wait a couple of more years before surgical robots take over doctors.
