A Real-Time Reconstruction-Tracking Synchronized Soft Tissue Deformation Evaluation Platform

Accurate quantification of intraoperative soft tissue deformations represents a critical challenge in computer-assisted surgery (CAS) systems. However, current methodologies lack consensus in achieving simultaneous submillimetric accuracy and real-time performance. In this paper, we establish a real-time reconstruction-tracking synchronized soft tissue deformation evaluation platform, addressing two main sub-problems: surgical instrument tracking and soft tissue surface 3D reconstruction. High-precision tracking of instruments is resolved through binocular near-infrared cameras, and high-resolution 3D surface patterns of the target tissues are captured through structured light reconstruction. Particularly, our integrated system achieves these two functions simultaneously by multiplexing the infrared cameras and designing the architecture of the simultaneous acquisition hardware. The standard phantom experiments indicate that our system enables submillimetric 3D reconstruction accuracy with a maximum point cloud reconstruction number of more than 1 million. The ex vivo experiments indicate close approximation to the ground truth CT-based evaluation, with MAE and RMSE of 0.521 ± 0.123 mm and 0.747 ± 0.227 mm, showing 45.0% and 63.4% improvements over the optical scanner. In addition, as a representative application scenario of our proposed system, we implemented our proposed platform in the real-time interactive evaluation of a typical biomechanical simulation model, achieving an update rate of 120 Hz (tracking)/10 Hz (reconstruction). The qualitative and quantitative findings indicated our proposed platform represents a significant step toward establishing a universal standard for evaluating soft tissue deformation, with promising potential to enhance the clinical application in computer-assisted surgery.