A surgical robot developed by researchers at Johns Hopkins University has autonomously performed a key portion of a gallbladder removal procedure without human intervention, marking a new step in the advancement of surgical robotics.

The system, named Surgical Robot Transformer-Hierarchy (SRT-H), was trained using videos of human surgeons operating on pig cadavers. During testing, the robot carried out a complex sequence of 17 surgical tasks; identifying ducts, placing clips, and cutting tissue — achieving results that researchers say were on par with expert surgeons. The work is detailed in a study published in Science Robotics.

Unlike earlier systems that followed rigid, preprogrammed plans in controlled environments, SRT-H responded dynamically to surgical scenarios. It adjusted to changes in anatomy, visual cues altered with blood-like dyes, and even changes in its initial positioning. Researchers also tested its ability to respond to voice commands, such as “grab the gallbladder head,” and corrections like “move the left arm a bit to the left.”
“This advancement moves us from robots that can execute specific surgical tasks to robots that truly understand surgical procedures,” said Axel Krieger, a medical roboticist at Johns Hopkins. “This is a critical distinction that brings us significantly closer to clinically viable autonomous surgical systems.”

The robot’s underlying AI architecture is similar to the models that power systems like ChatGPT. According to lead author Ji Woong "Brian" Kim, now at Stanford University, this interactive approach — combining video training with real-time feedback — addresses several barriers that have limited robotic autonomy in surgery.

SRT-H builds on prior work by Krieger’s team, including the Smart Tissue Autonomous Robot (STAR), which, in 2022, performed an autonomous laparoscopic procedure on a live animal. However, that earlier system required extensive environmental controls and precisely labeled tissue.

The project received funding from ARPA-H, the National Science Foundation, and the National Institutes of Health. Researchers plan to expand testing to additional procedures with the long-term goal of enabling complete autonomous surgeries.

Artificial Intelligence Homepage