A robot trained using surgical videos has performed a gallbladder removal on a life-like patient model, achieving a level of precision described as equal to that of a skilled human surgeon.
The system completed 17 tasks independently, including identifying blood vessels, placing surgical clips, and cutting tissue with scissors. It also adapted to changes during the procedure, such as altered tissue appearance caused by dye.
Unlike earlier surgical robots that followed pre-programmed steps or relied on specially marked tissue, this system could respond to spoken instructions and adjust its actions in real time.
Developed by researchers at Johns Hopkins University in the US, the robot uses a machine learning model based on the same architecture that powers ChatGPT. It was trained by watching surgical videos rather than following rigid, scripted commands.
During the operation, the robot identified ducts and arteries – the tubes that carry bile and blood – grasped them precisely, applied clips to prevent bleeding, and cut tissue using surgical scissors.
Axel Krieger, associate professor of mechanical engineering at Johns Hopkins, said: “This advancement moves us from robots that can execute specific surgical tasks to robots that truly understand surgical procedures.”
“This is a critical distinction that brings us significantly closer to clinically viable autonomous surgical systems that can work in the messy, unpredictable reality of actual patient care.”
The work builds on a 2022 project in which a robot performed the first autonomous robotic surgery on a pig. That system required specially marked tissue and operated in a controlled environment with a pre-set surgical plan.
Krieger said that earlier phase was “like teaching a robot to drive along a carefully mapped route”.
The latest model, he added, is “like teaching a robot to navigate any road, in any condition, responding intelligently to whatever it encounters.”
“To me, it really shows that it’s possible to perform complex surgical procedures autonomously.”
The robot responded to voice commands such as “grab the gallbladder head” or “move the left arm a bit to the left” and adapted its actions based on feedback.
The system also adjusted to individual anatomical features in real time and could correct its own performance as it progressed.
Ji Woong Kim, a former postdoctoral researcher at Johns Hopkins and now at Stanford University, said: “Our work shows that AI models can be made reliable enough for surgical autonomy – something that once felt far-off but is now demonstrably viable.”
Although the robot took longer than a human surgeon to complete the procedure, the researchers said its results were comparable to those of an expert.
The team plans to train and test the system on other types of surgery. The research was supported by US federal government funding.
Nuha Yassin, consultant colorectal surgeon and council member and lead for the future of surgery, robotics, and digital surgery at the Royal College of Surgeons of England, said: “It’s always exciting to see new developments in surgical innovation, especially in areas like robotics and digital surgery.”
“That said, the real test will be how safely and effectively the findings of this study can be translated into human trials. We need to make sure that progress doesn’t come at the expense of patient safety.”
