Robots operating on people has just taken a step closer to becoming a reality.
Researchers led by Johns Hopkins University revealed this week that they have trained a robot to perform surgical tasks by feeding it videos. The results have been successful, with it performing just “as skillfully as the human doctors”.
The authors say using tech like this will be able to cut down on medical mistakes and increase accuracy during surgery. They also add that this is one step closer to robots being able to carry out full surgery without human help.
“The successful use of imitation learning to train surgical robots eliminates the need to program robots with each individual move required during a medical procedure,” said the press release from Johns Hopkins.
How was the robot trained in surgery?
Along with researchers from Stanford, the team trained the da Vinci Surgical System robot via imitation learning. It was taught three key tasks: using a needle, lifting tissue and stitching.
The robot was fed hundreds of videos taken by da Vinci robots during surgeries. “We find that even with a few hundred demos, the model is able to learn the procedure and generalize new environments it hasn’t encountered,” said lead author Ji Woong “Brian” Kim.
In every task, the robot could perform just as well as a human doctor.
“The model is so good learning things we haven’t taught it,” said Axel Krieger, Johns Hopkins’ assistant professor in mechanical engineering.
“Like if it drops the needle, it will automatically pick it up and continue. This isn’t something I taught it do.”
Krieger described the development as “magical”, adding: “We believe this marks a significant step forward toward a new frontier in medical robotics.”
He said that previously programmers had to hand-code every line for surgical procedures, with modeling something like stitching potentially taking a decade.
The next step for the team is to train a robot to carry out full surgeries with imitation learning.
Feature image credit: Johns Hopkins University
