Why An Autonomous Robot Won’t Replace Your Surgeon Anytime Soon
Earlier this year, robots withdrew in defeat from the hospital.
Johnson & Johnson had been trying—operative word, trying—to sell a robot that can put patients to sleep for simple procedures like colonoscopies. Who did not like this? Human anesthesiologists, of course, whose jobs would be on the line. Professional groups lobbied hard against the device, questioning its safety. Even after the Food and Drug Administration approved the device, hospitals were wary. In March, Johnson & Johnson quietly stopped selling its anesthesiology robot.
But the robots are coming—they’re just not coming for any doctors’ jobs yet. To get hospitals to trust robots, their makers have realized, you can’t make doctors obsolete—you have to make them feel like Iron Man. Hence the popularity of Intuitive Surgical’s da Vinci system for minimally invasive surgery, whose insect-like arms are operated by a surgeon sitting at a control booth.
Another tack is automating repetitive tasks that can stretch surgeries into 16-hour marathons. Orthopedic surgeons already use automation to assist with their literally bone-crushing surgeries. But a new device, unveiled today by surgeons and engineers at Children’s National Medical Center and Johns Hopkins, could automate surgical tasks on floppy soft tissue. In a proof of concept, the team had the machine suture together two ends of a pig’s bowel.
The Smart Tissue Autonomous Robot could sew more evenly and consistently than even an experienced surgeon, according the report published in Science Translational Medicine. “It is a really nice piece of work. They’ve managed to push the envelope” says Ken Goldberg, director of UC Berkeley’s Center for Automation and Learning for Medical Robotics, who was not involved with STAR.
But in this case, STAR was still dependent on a surgeon to make the initial incision, take out the bowel, and line up the pieces before it fired up its autonomous suturing algorithm. “When you drive a car you use cruise control. The same logic would apply for surgical technology,” says Peter Kim, a pediatric surgeon on the STAR team. Just as cars have gained more autonomous functions—parallel parking, lane changes—STAR has been programmed to do other things like cut and cauterize, and Kim says they’re planning to do an entire supervised surgery like removing the appendix. But unlike Google’s autonomous car, which doesn’t even have a steering wheel, nobody is talking about a surgery robot with no human supervision.
If the technology behind STAR is going to make it into the hospital any time soon, according to Kim, it’ll probably be integrated into an existing platform. That could mean, for example, adding automated tasks to something like da Vinci, where the doctor still has final control. It makes sense, because the real advance behind STAR is software, not hardware. The robotic arm is just an arm from the German company Kuka, which makes robotic arms of all sizes for industrial use. What makes STAR unique is its ability to “see” inside the 3-D folds of soft tissue by using a 3-D lightfield camera—similar in concept to Lytro’s camera—that looks for fluorescent biomarkers injected inside the tissue. “The key to this paper is smart imaging technologies,” says W. Douglas Boyd, who specializes in robot-assisted heart surgery at the UC Davis Health System. “This is where the huge leap of advancement in these autonomous systems will be.”
In theory, adding such a camera and porting the software over to an existing platform should not be too hard. So is anyone interested? “Definitely. Yes yes yes,” says Dennis Fowler, executive vice president at Titan Medical, which is making a surgical system to compete against the da Vinci.
But technical capability isn’t the only barrier to acceptance among surgeons. Mazor Robotics makes a robotic system that identifies where surgeons should insert bone screws into the spine. Their machine could have easily done the drilling too, but Mazor found that surgeons preferred to give the go ahead and hold the drill themselves. “We had the technical ability to do it, but you have to go one bit at a time,” says Christopher Prentice, Mazor’s CEO. “The key to robotics in surgery is to add value, and I believe it’ll be incremental value. It’s not someone who swoops in.” That was Johnson & Johnson’s mistake with its anesthesiology bot.
It’s the same with cars. No one is trying to sell you a fully autonomous car yet. But the cars we do drive are already becoming increasingly automated, first with cruise control and now with lane change and parking assist. You’ll be lulled into trusting the robot driver and robot surgeon. But lulling will be slow and incremental. You can’t put the cart before the robot.