Leveraging augmented reality for dynamic guidance in 3-dimensional laparoscopic and robotic liver surgery: a prospective case series study.

Abstract

Purpose: Accurate anatomical knowledge and precise visualization are critical during liver surgery. We developed augmented reality (AR) software that overlays digital 3-dimensional (3D) models onto laparoscopic or robotic views, providing real-time visual aids for surgical navigation during 3D laparoscopic and robotic liver surgeries. This study assesses the accuracy of manual registration and the subjective perception of this AR software by the operator.

Methods: Ten consecutive patients undergoing 3D laparoscopic or robotic liver surgery from December 2023 to February 2024 were selected for application of the AR software during surgery. Manual registration accuracy was quantified post-registration using the Dice similarity coefficient (DSC) to compare the stereoscopic and virtual liver images. A 6-question operator survey, using a 5-point Likert scale, was administered after each surgery to evaluate the software's helpfulness in clinical settings.

Results: Seven males and 3 females (mean age, 62.4 ± 6.4 years) underwent liver surgery (3D laparoscopic, 5; robotic, 5). Surgical procedures included 4 right hemihepatectomies, 1 extended left hemihepatectomy, 1 left lateral sectionectomy, and 4 segmentectomies. The mean tumor size was 4.4 ± 2.2 cm (range, 1.0-7.5 cm). The mean DSC was 0.912 ± 0.052 (range, 0.879-0.954). The operator rated registration alignment favorably before (mean score, 3.9 ± 1.1) and after mobilization (mean score, 4.1 ± 1.2). The software was reported as very helpful overall (mean score, 4.2 ± 0.8), and in locating blood vessels (4.2 ± 0.6) and tumors (4.3 ± 0.7).

Conclusion: Clinical application of the AR software during 3D laparoscopic and robotic liver surgery is feasible, with favorable registration accuracy and high operator perception of helpfulness.