Performance of robot exceeds expert surgeons in preclinical models
What if your next surgery was planned and performed by a robot? A team at Johns Hopkins University is working to turn this idea into reality.
The concept of robot-assisted surgery is not new: several systems have already been developed and are being used to treat human patients. One example is the da Vinci surgical system, a laparoscopic device with robotic arms that are remotely controlled by a surgeon. This system is not autonomous—the robot does not perform any surgical tasks independently. Other robotic systems with higher levels of autonomy have been developed, such as the TSolution One®, which uses a robot to precisely cut bone according to a pre-specified plan. Existing autonomous robotic systems have largely been used to assist in surgeries involving hard tissues, such as drilling into bone for hip or knee implants. But these systems haven’t been used for soft tissue surgeries, which pose unique challenges, like accounting for unpredictable tissue motions that occur when the patient breathes, or size limitations of the surgical tools.
Now, NIBIB-funded researchers are developing an autonomous robot that can perform bowel surgery with minimal assistance from a surgeon. What's more, the robot outperformed expert surgeons when compared head-to-head in preclinical models. A study detailing the development of this robot, which showcases the first known autonomous laparoscopic soft tissue surgery, was recently published in Science Robotics.
“Surgical outcomes are highly dependent on a surgeon’s skill and experience, and even one missed stitch in a bowel surgery could lead to internal leak and infection,” said Moria Bittmann, Ph.D., program director in the division of Discovery Science & Technology at NIBIB. “This preclinical work is an important step towards autonomous robotic surgery in soft tissues, which could provide increased efficacy and safety in human patients, independent of the surgeon.”
The robot, called STAR (for Smart Tissue Autonomous Robot), was developed by Axel Krieger, Ph.D., and his colleagues at Johns Hopkins University. So far, the robot has been developed to perform intestinal anastomosis—where two pieces of small intestine are sewn together to form a single, continuous section—under the supervision and guidance of a surgeon. Krieger explained how the robot performs the procedure: After the surgeon manually exposes the tissue edges, STAR takes images and develops a plan for suture placement based on the shape and thickness of the tissue. Once the human operator approves of the plan, STAR independently stitches the tissue together. If the tissue deforms or moves beyond a set threshold, STAR asks the surgeon if a new surgical plan should be created. This process is repeated until the robot completes the entire procedure.
“By incorporating novel suturing tools, imaging systems, machine learning algorithms, and robotic controls, the STAR system is equipped to overcome the challenges of autonomous laparoscopic surgery in soft tissues,” said Krieger. “STAR can visualize a surgical scene, generate a surgical plan, and then execute those plans with high accuracy and precision.” He noted, however, that STAR is not meant to replace surgeons. “Autonomous robots, like STAR, are designed to be incorporated into the surgical workflow alongside surgeons, enhancing the performance of precise, repetitive tasks and ultimately improving surgical consistency from patient to patient.”
To evaluate how well STAR performed compared with expert surgeons, the researchers used “phantom” bowel tissues as a model system. Synthetic small intestine was mounted to a linear stage that was programmed to move back and forth, Krieger explained, which simulated breathing motions that would occur during surgery. Also during these experiments, the phantom tissue was randomly rotated and deformed, requiring STAR or the surgeon to pause, regroup, and complete the procedure, he said. STAR performed the procedure on phantom tissues five times, and four surgeons performed the procedure two different ways—twice using traditional manual laparoscopy, and twice using a different robot-assisted system.
When compared with the expert surgeons, STAR had fewer mistakes and was more consistent in suture spacing and depth. Additionally, when the researchers flowed viscous liquid through the resected phantom bowels, they found that the flow was the most laminar (smooth and streamlined) in the tissues reconstructed by STAR, indicating a higher quality anastomosis than those performed by the expert surgeons.
Finally, STAR’s performance was evaluated in a large animal model. Intestinal anastomosis was performed on five pigs; for four of the animals, the procedure was performed via STAR, and for the fifth animal, the procedure was performed via traditional manual laparoscopy. Similar to the phantom experiments, STAR made fewer mistakes compared with the expert surgeon. Additionally, when the researchers analyzed how well the resected bowels had healed seven days after surgery, there was no observable difference in wound healing between the two different surgical methods.
“Our results indicate that STAR is more consistent and accurate than expert surgeons when performing suturing tasks,” said Krieger. He noted that their findings demonstrate the potential for autonomous surgical robotics to democratize surgical care—which could lead to more predictable and consistent patient outcomes.
“While many may feel hesitant about having a machine perform a specialized task that is traditionally performed by a human, robotic systems have the potential to improve patient outcomes in medial settings,” said Krieger. “Just like the public has embraced the gradual influx of cruise control, lane assist, and self-parking features in automobiles—which will eventually lead to self-driving cars—I think we will see a similar progression in the field of medical robotics.”
This study was supported by two grants from NIBIB (R01EB020610 and R21EB024707).
Study reference: H. Saeidi, J.D. Opfermann, M. Kam, S. Wei, S. Leonard, M.H. Hsieh, J.U. Kang, A. Krieger. Autonomous robotic laparoscopic surgery for intestinal anastomosis. Science Robotics 2022 7 (62). DOI: 10.1126/scirobotics.abj2908