Crack-Free Multi-Resolution Dynamic Volume Refinement for Physics-Based Bone Drilling and Shaving in Arthroscopic Procedures

Arthroscopy is a minimally invasive procedure used by orthopedic surgeons to diagnose and treat joint injuries. It involves using a 2D fiber optic view from an arthroscope displayed on a monitor, requiring high dexterity and practice. This study has two main goals: to develop a virtual reality training platform called Virtual Rotator Cuff Arthroscopic Skill Trainer (ViRCAST) for diagnosing and treating rotator cuff tears, and to introduce a novel algorithm for simulating bone shaving and drilling in arthroscopic surgery. ViRCAST and its hardware components are described, and realistic haptic interaction is achieved using Dynamic Proximity Hierarchy’s multi-point collision detection. A dynamic resolution approach improves haptic force feedback by generating finer voxels based on proximity to the surgical tool. The bone drilling simulation ensures smooth feedback even with coarse-resolution voxels. With a voxel size of 1.2mm, execution time is 6 ms, while reducing the voxel size to 0.54mm results in surface mesh creation in 34 ms and visual rendering at 29 FPS.