Real-time collision detection for long thin medical instruments in virtual reality-based simulators

Realistic haptic interaction represents an important requirement for successful virtual reality-based simulators that could replace traditional surgery training methods. This paper presents a new collision detection algorithm for long thin object used in hard tissue removal operations, e.g., sawing. The algorithm works efficiently by reducing the number of bone volume elements that needs to be checked for collision with the tool and also by reducing the number of transformations between the virtual scene coordinate systems. The algorithm is integrated into a Virtual Reality-based sawing training system for orthognathic surgery.