Investigating the role of simulation fidelity in laparascopic surgical training

Abstract

Minimally invasive surgery (MIS), with its aptitude for quick recovery and minimal scarring, has revolutionized surgery over the past few years. As a result, the development of a VR-based surgical trainer for MIS has been a popular area of research. However, there still remains a fundamental question of how realistic the simulation has to be for effective training. On the one hand, learning surgical practices with an unrealistic model may lead to negative training transfer. However, because of the learning abilities and perceptual limitations of the sensory, motor, and cognitive system of the human user, perfect simulation is unnecessary. Furthermore, given the large variations in human anatomy and physiology, there is no single perfect model. The question is how simple a simulation can we get away with, while at the same time preserving a level of fidelity between the virtual and real organ behavior that leads to positive training transfer. A dual station experimental platform was set up for this study. The two stations consisted of a real environment testing station and a virtual environment training station. The fidelity of the simulation could easily be adjusted in the virtual training station so that subjects could be treated with different modes of training. With the dual station setup the real environment performance of a subject before and after VE training could be measured.

(cont.) First round of experiments on the setup were conducted to investigate the effect of haptic fidelity and the effect of part task training on surgical training. Haptic fidelity was adjusted by modeling a material of non-linear stiffness to different degrees of accuracy. Subjects were initially tested on the real station performing a bimanual pushing and cutting task. They were then trained on the virtual station, with one of the three different levels of haptic fidelity or the part task trainer. Once the training was complete, the subjects were again evaluated on the real environment station to gauge their improvement in skill level. Initial results showed a marked difference in level of skill improvement between training with haptics and without. However there was no significance difference in the training effectiveness of the higher fidelity and lower fidelity model of elasticity. Also part task training proved to be an equally effective method of training for the surgical task chosen. Experiments with modeling the non-linearity materials are one of many studies that can be done on this platform, including adjusting other modes of haptic fidelity such as visco-elasticity and experiments with graphic fidelity. Results from such experiments can serve as the basis of future surgical simulation development by providing guidelines on environment fidelity required for positive training transfer to occur.