For the first time a robot he managed to perform a surgery on a simulator patient without resorting to any type of help from humans. Trained with videos of surgerythe robot SRT-H (Surgical Robot Transformer-Agerchy) was able to remove a gallbladder adapting in time real to the anatomical characteristics of the fake patient, making decisions on the fly And self -portraying itself When things did not go as expected. The result is published in the magazine Science Robotics by a research group led by Axel Krieger of Johns Hopkins University (USA).
The same team had already developed Another surgeon robot called Star (Smart Tissue Autonomous Robot) who in 2022 had performed the first intervention autonomous robotic surgical on a lively animal: in that case, however, it was a procedure in laparoscopy And the robot had performed a rigid and predetermined surgical plan, moving between fabrics that had been specially marked to facilitate the job.
“It was like teaching a robot to drive along a carefully mapped path,” explains Krieger. With the new system, however, “It’s like teaching a robot to travel any roadin any conditionresponding intelligently to anything you meet. This progress brings us from robot capable of performing specific surgical tasks a robot That they really include the surgical procedures. It is a fundamental distinction that brings us significantly to clinically valid autonomous surgical systemsable to operate in the chaotic and unpredictable reality actual assistance to the patient “.
Built with the Same learning architecture automatic Of ChatgptSRT-H is also interactive: during the intervention responds to voice commands (how ‘he grabs the head of the gallbladder’) and follows the suggestions (For example, ‘move your arm slightly to the left’), just like a novice surgeon who works with a mentor.
The cystifellea removal procedure performed in the study provides for a sequence of 17 tasks lasting a few minutes. The robot had to identify certain learned and arteriesgrab them precisely, strategically place the clips e tall The parts with scissors. SRT-H learned to perform every action watching Johns Hopkins surgeons videos that made pork corpses. Visual training was further strengthened with captions describing the activities.
After watching the videos, the robot performed the intervention with aaccuracy of 100%: During the various tasks he acted in an imperturbable way and with the competence of an expert human surgeon, even during unexpected scenarios typical of real medical emergencies. Although the robot took more time than a human surgeon to perform the intervention, the results were comparable to those of an expert surgeon.
“Our work – observes the first author of the study, Ji Woong ‘Brian’ Kim – shows that artificial intelligence models can be made enough reliable for autonomy in surgery, a goal that once seemed far away, but which is now clearly achievable”.
Reserved reproduction © Copyright Ansa
Source: Ansa
I am a journalist with over 6 years of experience working in the news industry. I currently work as an author for Global Happenings, and my coverage focuses on Technology news. I have written for various publications, including Reuters, The New York Times and The Guardian.
