Robustness and complexity of a minimally invasive vascular intervention simulation system

Abstract

Minimally invasive vascular interventions offer advantages over open surgery. Thorough training is needed to master the skills required to correctly perform these minimally invasive interventions. Simulation is becoming a potential alternative for training. We have developed the foundation, i.e., the algorithmics and models, for a minimally invasive vascular intervention simulation system focusing on guide-wire manipulations. In this article we address the robustness, accuracy, and complexity of this foundation using a phantom. To this end the theory on which the simulation is based is used to formulate constraints on the simulation parameters. Furthermore, the parameter space has been explored in order to optimize the trade-off we are facing: accuracy versus speed of the simulation. A physical experiment setup has been designed that allows us to obtain ground-truth data. The accuracy of the simulation has been determined by comparing physical experimental results with various simulations. For multiple combinations of parameter settings the simulation supplies a guide-wire configuration with a root-mean-square error around 1 mm. The results show that the speed of the simulation is still an issue and needs to be improved. Nevertheless, the results also indicate that the developed algorithms and models are a robust foundation to build a simulation system on.