Artificial intelligence-enhanced navigation for nerve recognition and surgical education in laparoscopic colorectal surgery

BackgroundDevices that help educate young doctors and enable safe, minimally invasive surgery are needed. Eureka is a surgical artificial intelligence (AI) system that can intraoperatively highlight loose connective tissues (LCTs) in the dissected layers and nerves in the surgical field displayed on a monitor. In this study, we examined whether AI navigation (AIN) with Eureka can assist trainees in recognizing nerves during colorectal surgery.MethodsIn left-sided colorectal surgery (n = 51, between July 2023 and February 2024), Eureka was connected to the laparoscopic system side by side, and the nerve was highlighted on the monitor during the surgery. We examined the rate of failure to recognize nerves by trainee surgeons over a total of 101 scenarios after it was recognized intraoperatively by the supervising surgeon (certified by the Japanese Society of Endoscopic Surgery). We also examined the frequency of nerve recognition by the trainee physicians viewing the Eureka monitor when recognition was not possible (recognition assistance rate).ResultsThe nerve recognition failure rate and recognition assistance rate with AIN were as follows: right hypogastric nerve during sigmoid colon mobilization, 44/101 (43.6%) and 19/44 (43.2%); left hypogastric nerves during dissection of the dorsal rectum, 27/101 (26.7%) and 13/27 (48.1%); right lumbar splanchnic nerves, 32/101 (31.7%) and 29/32 (90.6%); left lumbar splanchnic nerves, 44/101 (43.6%) and 39/44 (88.6%); and pelvic visceral nerves during dissection of the dorsal rectum, 29/45 (64.4%) and 6/29 (20.7%), respectively.ConclusionAlthough the rate of recognition with assistance from AIN differed for the different nerves, this system can potentially assist in anatomic recognition, enhance surgical education, and contribute to nerve preservation.Trial registration: Improvement of AI navigation in minimally invasive surgery and examination of its intraoperative support and educational effectiveness. Research Ethics Committee of the Kawaguchi Municipal Medical Center (Saitama, Japan) approval number: 2022-27.ConclusionAlthough the rate of recognition with assistance from AIN differed for the different nerves, this system can potentially assist in anatomic recognition, enhance surgical education, and contribute to nerve preservation.Trial registration: Improvement of AI navigation in minimally invasive surgery and examination of its intraoperative support and educational effectiveness. Research Ethics Committee of the Kawaguchi Municipal Medical Center (Saitama, Japan) approval number: 2022-27.