Evaluating User Performance in Augmented and Virtual Reality for Neurosurgical Targeting Tasks
Abstract
This thesis addresses the pressing issue of surgical errors through the development and comparison of an application aimed at enhancing neurosurgical training using Augmented Reality (AR) and Virtual Reality (VR). Surgical errors, often preventable and unintentional, lead to significant mortality rates and are primarily attributed to technical mistakes. The conventional medical training approach, limited by resources and potential risks to patients, necessitates a safer yet effective alternative.
The study's primary objectives encompass three key areas. Firstly, it aims to evaluate the feasibility and user performance of an application targeting neuroanatomy tasks, particularly focusing on External Ventricular Drain (EVD) procedures. This evaluation involves testing users with varying levels of neuroanatomy knowledge to assess accuracy and performance. The hypothesis anticipates higher accuracy from experts but potentially slower speeds due to technological novelty, contrasting with faster novice performance but lower accuracy.
Secondly, the research aims to discern the superior modality for surgical training between AR and VR. By directly comparing these modalities, the study hypothesizes that the transparent, see-through nature of AR may facilitate higher user accuracy compared to VR, despite VR's advantage in depth perception.
Lastly, the study endeavors to compare various modern AR and VR devices and Software Development Kits (SDKs) to determine if certain devices exhibit superior performance. This assessment aims to ascertain whether investing in specialized AR devices like glasses provides significant advantages over mobile AR applications in terms of cost and usability.
The potential impact of this research lies in its ability to guide the design of future surgical training applications based on the superior technology and performance uncovered. If validated, this application could serve as a viable alternative for neurosurgical training without the need for cadavers. Additionally, it could inform the selection of technology for future applications based on performance metrics.