In the Magazine Tech & Science

Robot Surgeon Succeeds Without Help from Human Doctors

05_04_Robot_Surgeon
The Smart Tissue Autonomous Robot can perform soft tissue surgery. Dr. Azad Shademan and Ryan Decker are pictured during supervised autonomous in-vivo bowel anastomosis performed by the Smart Tissue Autonomous Robot (STAR). Axel Krieger/AAAS

Humans make mistakes. Even surgeons, who train intensively for years and adhere to exacting protocols, are not infallible. But what if these doctors could pool their knowledge and experiences together and create an optimal surgical standard of care, to be carried out by machines? Computers, after all, make precise calculations effortlessly and consistent motions without tiring. 

That’s the idea behind surgical robots, which may soon perform most surgeries, from sewing up tiny wounds to executing heart procedures. Many of these operations are, in fact, already accomplished with the assistance of robots, with machines like the da Vinci Surgical System mimicking every move a surgeon makes as she controls the tools from afar. But a recent test conducted by researchers at Children’s National Health System (CNHS) and Johns Hopkins University suggests that robots in the operating room may soon go a step further, performing on soft tissue completely on their own, from start to finish. The surgeons, meanwhile, would simply stand by and watch.

The Smart Tissue Autonomous Robot, or simply STAR, successfully completed open bowel surgery in pigs, leaving the animals healthy and without complications.

“We’re the first group to [develop autonomous robotic surgery] with soft-tissue surgery, and when compared to standard practice, it’s better,” says Peter Kim, professor of surgery and vice president of the CNHS Sheikh Zayed Institute for Pediatric Surgical Innovation. “The idea is not to replace surgeons; it will make the surgeons better and make the procedures safer.”

A recent Mayo Clinic study found that major surgical errors—including operating on the wrong site or side of the body, inserting the wrong implant, or even leaving tools or objects inside the patient—occur every one out of 22,000 invasive procedures. That’s rare, but robots like STAR would aim to lower the number even further.

The motivation behind STAR began years ago, on top of a skiing hill in Canada. Back then, Kim was a pediatric reconstructive surgeon, an expert in performing airway reconstructive surgeries on tiny babies. While skiing with his son one winter, he received a call about a baby across the country needing immediate surgery. Rushing down the hill, he had a realization: “I didn’t get down safely,” he says, “it wasn’t good for me, and it wasn’t good for the patient.” That got him thinking about the possibility of transferring all of his surgical knowledge to one place—a computer program, stored in a machine that colleagues could access it at anytime, anywhere. In theory, it would contain an accumulation of master surgeon experiences, complementing humans in the operating room and holding standard care to the highest level. “Why wouldn’t I want to have technology that could make me a better surgeon?” he says.

Robot-assisted surgeries have been around for nearly two decades. In the da Vinci surgical system, approved for clinical use by the Food and Drug Administration in 2000, surgeons place their arms inside instruments and use their hands to control the movement of robotic tools on the operating table from afar. Though this was designed to improve to extend the dexterity of a surgeon’s hand, some debate remains on whether it’s truly that much more effective than standard surgery, and the practice certainly had “room for improvement,” Kim notes. It’s what Kim refers to as “master-slave” machinery, where the robot’s every major move is manually controlled by surgeons, and thus its results may vary based on the surgeon’s training or experience.

STAR, on the other hand, is entirely autonomous. It’s not only capable of working on its own and performing surgical motions with a more flexible “hand,” but it’s intelligent and able to react to the uncertain, dynamic landscapes of soft tissue surgery without direction. Cutting into hard tissue like bone is one thing, but operating on moving soft tissue—like the “different shades of pink” of the abdomen, Kim says—is far more complicated. STAR reacts to a changing environment, similar to how self-driving cars are programmed to not only drive on the highway, but also to react to someone else making a mistake and getting in your lane.

The other unique feature of STAR is its “suprahuman” visual system, which consists of multiple lenses that tell the robot where objects are in three-dimensional space, then add in other capabilities like seeing through tissue or seeing through darkness, like night vision goggles. The visual system can also tell what type of tissue the robot is looking at, its temperature, and whether it has a natural rhythm, like a heartbeat.

To test the robot, the researchers conducted several open bowel surgeries—or intestinal anastomosis—on pigs. All of the surgeries were successful and the pigs experienced no complications. When they compared the results to those of manual surgeries, laparoscopies and robot-assisted surgeries by the da Vinci Surgical System, they found that STAR actually performed better, with fewer errors and more consistent sutures.

As for the surgeons standing guard? “Sixty percent of [the surgeries] were done fully autonomously and we made minor adjustments in 40 percent of cases—not because robot needed it,” Kim says, “but because we were like expectant parents watching the child walk for the first time, so we were a little nervous.”