Skip to Main Content

Its name is STAR. And on Wednesday, scientists reported that the Smart Tissue Autonomous Robot could stitch together separate pieces of the bowel in pigs, the first time a surgical robot has completed a portion of an operation in living soft tissue without human guidance.

The new paper, published in the journal Science Translational Medicine, amounts to what’s known as a proof of concept — a demonstration that the new advance appears to be possible. Outside experts described it as a key achievement in efforts to move toward autonomous robotic surgery, but noted the technology is years away from being used in operating rooms, or even in a clinical trial.


“It’s one step forward,” said Dr. Dragan Golijanin, director of the Minimally Invasive Urology Institute at the Miriam Hospital, an affiliate of Brown University, who was not involved in the research.

Surgeons use robots in operating theaters around the country, but as it stands, they guide them like a puppeteer directs a marionette, conducting every move and response.

A number of studies looking at different operations have found that these robot-assisted surgeries have failed to improve outcomes over standard surgical methods, but backers of the technology say clinicians more experienced with the devices do have better results.


The goal of robotic surgery is not to replace surgeons, supporters of the method emphasize. And the system described in the new paper would be supervised by a surgeon who could step in in the case of an emergency.

But if robots could perform a portion of operations on their own, boosters say, it could reduce complications caused by human error, speed up surgeries, open up operating room availability, and enable patients to have the best techniques available to them no matter where they live.

“Consistency in surgery is important, just like in manufacturing,” said roboticist Ken Goldberg, codirector of the Center for Automation and Learning for Medical Robotics at the University of California, Berkeley.

Soft tissue presents a particular challenge for autonomous robotic surgery. Instead of sawing through bone, for example, robots trying to suture squishy organs need to differentiate between different tissues and adjust as the tissues change shape.

“It’s hard to manipulate even rigid objects — it’s hard for a robot to clear a dinner table,” said Goldberg, who was not involved in the research published Wednesday.

To develop STAR, the researchers started with an existing robot arm from KUKA Robotics and introduced two imaging innovations to overcome the problems imposed by operating on soft tissue. They enabled the robot to track shifting shapes in 3-D and added a function, described almost like night vision, that helped the robot tease apart which tissue to suture and which to leave alone.

The researchers have filed about half a dozen patent applications for the technology and hope to see it commercialized in the future. They said the soft-tissue surgeries that could first be done autonomously include removals of the gallbladder and appendix.

“Having a tool like this, and by making the procedures more intelligent, we can have better outcomes for patients,” said Dr. Peter Kim, a pediatric surgeon at Children’s National Health System and senior author on the paper.

With their system, the researchers from Children’s National and Johns Hopkins University programmed the robot with skills taken from top surgeons, they said. An algorithm instructed STAR where it should insert the needle during sutures, how much space should be left between each stitch, and how tight to pull the suture.

The team compared the system to existing stitching methods including robot-assisted sutures, laparoscopic sutures, and hand-sewn sutures.

In one test in four live pigs, STAR made just a few suturing mistakes in connecting pieces of bowel — a procedure called intestinal anastomoses — and it took several times as long as a human surgeon, who made zero mistakes sewing sutures by hand in the one control animal. But the researchers said that they were not trying to see how fast the robot could perform the procedure, just that it could do it.

About 60 percent of the time in the experiments, the researchers left the robot to do what it would, and they made sure STAR was placing the sutures in the correct place the rest of the time. The researchers, Kim said, were like parents watching their babies learn to walk, encouraging them to try it on their own but there to ensure the steps were being taken as they should.