Augmented reality during rectal cancer surgery: Feasibility study and preliminary experience

Aim To enhance laparoscopic rectal cancer surgery, we implemented Augmented Reality (AR). Methods Magnetic resonance neurography and CT images were segmented to create 3D pelvic models, including rectum, nerves, vessels, and ureters. The use of Artificial intelligence with ConvNext U-Net binary segmentation architecture addressed real-time occlusion of surgical instruments in AR overlays. Anatomical landmarks such as the abdominal aorta bifurcation and the right iliac artery crossing the ureter were used to anchor the 3D model to the surgical view. Retroperitoneal structures were preferred for AR application due to their stability under pneumoperitoneum. Results Key high-risk points during anterior rectum resection were identified to focus AR use on the most critical surgical phases. Ten accurate and reproducible 3D pelvic models, including tumors, were developed. AR-assisted surgery was successfully conducted in two laparosopic rectal resection cases. During these procedures, surgeons' ability to identify essential pelvic structures-vessels, ureters, and nerves-was assessed. Conclusion The study demonstrates the feasibility of AR in rectal surgery and suggests that it may enhance surgical precision and safety by improving intraoperative visualization of complex pelvic anatomy.