Mobile Augmented Reality for Craniotomy Planning

Abstract

Augmented reality (AR) neuronavigation has been proposed to address the shortcomings of conventional neuron-avigators. Researchers have presented low-cost AR methods for craniotomy planning, but they lack navigation capabilities. Other studies introduce AR neuronavigation systems that are a step further in usability than traditional neuronavigators, but they may be hard to obtain or reproduce. AR neuronavigation was also implemented on mobile devices, but most systems have an undesired lag during the navigation. This work investigates the feasibility of creating an accurate and low-cost standalone mobile AR neuronavigator. Unlike other mobile approaches, this solution has no perceptible lag as the processing is efficiently performed on the device instead of an external computer. Results show that a neuronavigation system can be deployed on a mobile device, running smoothly at 60 frames per second, and achieving a smaller than 5 mm target registration error.