With only a small amount of guidance, a robotic surgeon was able to stitch together pieces of intestinal tubing from a pig — in both a lab setting and an operation on a living animal. This type of procedure is particularly difficult as the soft tissue can bend and flex, but it’s necessary to perform a number of life-saving operations. The researchers claim their robot can match and even exceed the safety and precision of a human doctor.

In the study published in Science Translational Medicine, the researchers caution that their new method is only a proof-of-concept and requires more development. They’ll need to provide evidence their techniques are safe enough for people, and even there they may face difficulty. Robotic surgery techniques that have been used in hospitals for years are still criticized as no safer than non-robotic methods, but much more expensive. The researchers say that in the future there will be all sorts of ways for humans and robots to work together to benefit patient safety, but that the end goal is a machine that operates entirely by itself

Of course, it’s not ready for primetime just yet. "It’s a very provocative idea and I think it has great potential," Rasa Zarnegar, an associate professor of surgery at Weill Cornell Medical College told The Verge. "But the study is very small [and] definitely doesn’t portray a real-life scenario."

STAR in action, getting to grips with some pig intestinal tissue. (Image Credit: Carla Schaffer / AAAS)

The key to the automated surgeon is advanced vision, say the researchers that developed it from Children’s National and John Hopkins University. Like the facial tracking technology used in Hollywood films, the robot — called Smart Tissue Autonomous Robot or STAR — uses markers to follow the movement of flesh. In this case, these are fluorescent dots injected into the pig’s tissue that are tracked by an array of cameras and mapped onto a 3D model. Computer software that has been compiled using data from past surgeries (only "the best," says Kim) then decides where to place the stitches on this 3D model. This system allows the robot to track the tissue as it bends and flexes, and the trackers are visible even in the "very homogeneous, non-textured environment" of the human body. That is to say, the robot can see the markers through all the blood and gore of normal surgery. At the moment, though, the dots still have to be placed by a human operator.

The rest of the STAR system is built around a robotic arm from German industrial robot manufacturer Kuka. The arm is equipped with seven rotating axes and a suturing device fitted on the end. It’s like one of those grabber sticks used to pick up trash, but with a tiny curved needle at the bottom, which rotates like a spinning wheel to sew up the tissue. A force sensor provides feedback on how hard the needle is pushing, and a computer display tracks the surgery’s progress. And although the STAR system decides where to place each suture, a human surgeon has to approve each one, manually repositioning the stitch if necessary. The STAR system was about 60 percent autonomous in these tests, but the researchers say they were being overly cautious. "We were like parents with young child that was about to walk," said Kim. "We were looking at it carefully."

The STAR robot's skills were tested on intestinal material from a pig, sewing up a lengthwise cut in one experiment, and suturing together two pieces of tubing (a procedure known as anastomosis) in another. The first experiments were conducted in a lab setting with the tubing secured in a harness, while the second was repeated using a live, anesthetized pig. The robot was judged on the quality of its stitches and its ability to produce a leakproof intestinal tubing, then compared with three other human-based methods. First, a human doctor using regularly open surgery tools; second, using keyhole surgery tools; and third, using the da Vinci — a mechanical system controlled completely by a human and used in the vast majority of robot surgeries. In each case, the tissue was prepared by a third party for suturing.

STAR getting a helping hand from a supervisor. (Image Credit: Carla Schaffer / AAAS)

Overall, the STAR robot produced stitches that were more evenly spaced than any of the other three methods. This is a good signifier of a suture’s strength, said Axel Krieger, an assistant to the study, as inconsistent spacing leads to pressure points. These, in turn, lead to leaks. "The procedure is like trying to put together a garden hose which has been cut," said Krieger during a press call. "[You] want to be strong enough so that when it increases in pressure, the garden hose will not leak."

In the first experiment, the STAR system was able to track the target tissue as it was moved about between stitches to mimic real surgery. The robot’s stitch placement also needed to be corrected less than both the da Vinci and keyhole surgery methods. On this measure, it did about as well as the human performing regular open surgery. When it came to comparing the strength of the stitched intestines (those leaky garden hoses), the STAR system’s finished tubes were as good as the human surgeon’s, able withstand the same or equal pressure in tests. The main drawback in the final test was that the system was significantly slower than a human, although its creators again attribute this fact to their own caution. "We can run the robot really, really fast," said Krieger. "But in this study we really focused on the mission and the outcome, so we didn’t run it as fast as we could."

The STAR system isn’t ready to don scrubs and enter the ER, though. The experiments conducted were limited in scope and number (only four in the last test, compared to only one control), and performed under tightly-controlled conditions. Zarnegar pointed out that that failure rate for these sorts of surgical procedures is less than 1 percent when not using robotic surgeons, so the study’s sample size doesn’t really reflect potential dangers. "I think you need bigger studies to prove that automated systems can do it better," says Zarenagar. Another expert in robotic surgery, Naeem Soomro, told The Verge that the paper was was "very high quality," but that it might take another 10 years before such systems are used in surgeries. "But ultimately it is all about patient safety, if we can do better operations with fewer complications, we should."

Dr Shadman and Ryan Decker during supervised in vitro bowel anastomosis. (Image Credit: Axel Krieger)

Past forms of robot surgery, though, have been less straightforwardly beneficial. Studies have suggested that the da Vinci system, which was cleared for use by the FDA in 2000 and is by far the market leader, has a similar complication rate to traditional surgery for certain procedures, despite the greater expense. (With add-ons, an individual system can cost in excess of $2 million.) Patients have also sued the robot’s manufacturer, Intuitive Surgical, for allegedly failing to adequately train doctors to operate the da Vinci.

Arnold Advincula, a specialist in minimally invasive surgery, said that incidents like these may have made it more difficult for the STAR system to get quick approval. "In some ways, today’s robotic platforms left a little bit of a bad taste in our regulatory agencies," he told The Verge, noting that one medical robot company, TransEnterix, recently had its surgical bot rejected by the Food and Drug Administration. "The FDA is not just handing these things out anymore," says Advincula. He still believes, though, that automated surgical procedures will be folded into regular surgical practice in the future — the robots won’t be the only surgeons, but they'll certainly be a part of the operation.

The creators of STAR are confident, too. They say that their use of existing technologies like the Kuka robotic arm mean their robot can be adopted more quickly for clinical use. They also hold several patents for the STAR system, and suggest that in the future they may talk to "financially interested partners." The next step, though, is more case studies and experiments to find out the strengths and weaknesses of automated systems. Kim suggests that as the they’ve already tried sewing up tissue, the next challenge should be removing something. "Being able to demonstrate that you can do a critical procedure such as an appendectomy would be of huge value," says Kim.

The researchers compare the potential of robot surgery to self-driving cars. In both cases, we may be squeamish about the idea of relinquishing control to machines, but if the result is fewer human casualties, why complain? "The same logic would apply for surgical technology," says Kim. He adds that in the future, human surgeons may take orders from robots, or they may work alongside one another. Either way, he says, if it saves lives, it’s worth exploring.