Video-Based Soft Tissue Deformation Tracking for Laparoscopic Augmented Reality-Based Navigation in Kidney Surgery.

Area of Science:

• Medical Imaging
• Surgical Navigation
• Augmented Reality

Background:

• Minimally invasive surgery (MIS) presents challenges in tracking critical structures within dynamic patient anatomy.
• Accurate real-time navigation is crucial for improving outcomes in laparoscopic procedures, particularly kidney surgery.

Purpose of the Study:

• To develop and evaluate an augmented reality (AR) navigation pipeline for laparoscopic kidney surgery.
• To address challenges in initial registration and soft tissue deformation tracking using preoperative imaging and laparoscopic data.

Main Methods:

• Utilized preoperative MR/CT data and binocular laparoscopes for registration and deformation tracking.
• Employed a probabilistic rigid registration algorithm with elastic compensation and dense point cloud reconstruction.
• Integrated a deep learning-based feature point selection strategy with a biomechanical model for continuous deformation tracking.

Main Results:

• Ex-vivo experiments showed external surface reprojection error of 4.07 ± 2.17 mm and internal structure error of 2.98 mm.
• In-vivo experiments demonstrated mean absolute errors of 3.28 ± 0.40 mm and 1.90 ± 0.24 mm for internal structures.
• The system achieved accurate tracking of soft tissue deformation without additional interventional hardware.

Conclusions:

• The proposed AR-assisted navigation system shows significant potential for enhancing clinical applications in laparoscopic kidney surgery.
• The pipeline effectively addresses registration and deformation tracking challenges in MIS.
• Quantitative and qualitative results support the system's accuracy and utility in surgical guidance.