Real-time precision with augmented reality in liver surgery

Abstract Introduction Liver surgery is associated with significant clinical challenges due to complex anatomy and the risk of complications. Enhanced intraoperative navigation tools may improve surgical outcomes by increasing resection radicality while minimizing adverse events. This preclinical proof-of-concept study evaluated a novel augmented reality (AR) system for intraoperative guidance in liver tumor surgery. Method A tumor model was implanted into the liver of porcine subjects, and a landmark device containing a visible marker was affixed to the liver surface. Computed tomography established the spatial relationship between the tumor and the landmark. During surgery, the endoscopic camera detected the marker, generating a three-dimensional reconstruction that enabled continuous tumor tracking during resection. Resected specimens were assessed for radicality and AR accuracy data were collected. System usability was evaluated using the System Usability Scale (SUS). Result Six porcine underwent liver resection. Five procedures achieved radical margins, while one was non-radical. During the resection phase average AR accuracy was 3.4±2.0mm, with a mean landmark detection rate of 80%. Blood contamination affected visualization in approximately 5% of cases, while water had no impact. The mean SUS score was 67. No adverse events were observed. Discussion This novel AR system demonstrated high accuracy and tracking stability with promising usability outcomes. While radicality could not be achieved in all cases, this limitation was attributed to issues with the tumor model rather than the system itself. These results indicate potential clinical applicability with some aspects warranting further refinement.