A surgical robot has successfully executed a complex phase of gallbladder removal without human intervention, marking a significant milestone in robotic surgery, according to a new study.
The robot, trained on surgical videos, operated on a lifelike model and responded to voice commands from the surgical team, akin to a novice surgeon guided by a mentor.
The development, led by researchers at Johns Hopkins University and funded by federal grants, showcases the potential for robots to combine mechanical precision with human-like adaptability in surgical settings. The findings are detailed in the latest issue of Science Robotics.
Medical roboticist Axel Krieger 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.”
In a previous milestone, Krieger’s Smart Tissue Autonomous Robot (STAR) performed autonomous laparoscopic surgery on a live pig in 2022. This earlier model operated under controlled conditions, requiring specially marked tissue and following a predetermined surgical plan.
Krieger likened it to teaching a robot to drive on a mapped route. The new system, however, is designed to navigate varied conditions, responding intelligently to unforeseen challenges.
The Hierarchical Surgical Robot Transformer (SRT-H) represents a significant advancement. It performs surgeries by adapting to individual anatomical features in real-time, making immediate decisions, and self-correcting as necessary.
Built on the same machine learning architecture as ChatGPT, SRT-H can interact with the surgical team, responding to spoken commands and learning from feedback.
Lead author Ji Woong said: “This work represents a major leap from prior efforts because it tackles some of the fundamental barriers to deploying autonomous surgical robots in the real world.
“Brian” Kim, a former postdoctoral researcher at Johns Hopkins who’s now with Stanford University. “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.”
Previously, Krieger’s team trained the robot to perform basic surgical tasks such as needle manipulation, tissue lifting, and suturing, each completed in mere seconds. The gallbladder removal procedure, however, involves a more complex sequence of 17 tasks, requiring the identification and precise handling of ducts and arteries, strategic clip placement, and tissue severance.
SRT-H learned the gallbladder procedure by watching videos of Johns Hopkins surgeons operating on pig cadavers, supplemented by task-descriptive captions. Following this training, the robot achieved 100% accuracy in performing the surgery. Although the procedure took longer than it would for a human surgeon, the accuracy and outcomes were comparable to those of an expert.
The robot demonstrated consistent performance across varying anatomical conditions and unexpected challenges, such as altered starting positions and the introduction of blood-like dyes that changed the appearance of the gallbladder and surrounding tissues.
