An Intelligent Guidance System for Computer-Guided Surgical Training

Abstract

Laparoscopic surgery performed by expert surgeons enables patients to reduce recovery time and pain. However, this surgical procedure is challenging due to restricted vision and depth perception issues. Various simulators have been developed for simulation-based laparoscopic surgery skills training. Technical realization methods of existing simulators are summarized to develop the next generation training systems by conducting systematic review. In this dissertation, a novel computer-assisted surgical trainer (CAST) is presented to provide the better training experience. CAST guides a trainee actively by providing visual, force, and audio guidance. An augmented reality overlay system was implemented first using a single web camera. Also, an instrument tracking system was designed to support the guidance based on kinematic equations with particle swarm optimization (PSO). For the force guidance system, self-adjusting fuzzy sliding mode controllers are proposed with virtual fixtures. To generate a training task, modeling specifications and a text-based task description method are presented as well. While performing a training task, CAST collects movement data, then objective assessment results are provided to a trainee using a fuzzy scoring system with achievable goal-based evaluation metrics. The initial test data indicate the feasibility and effectiveness of the proposed computer-guided training system.