E2200. Learn Imaging Anatomy Using Interactive Volume Rendering and Anatomy Model Viewing in Virtual Reality: An Inexpensive Mobile Implementation
Authors
Nick Maizlin;
University of Western Ontario
Rohit Malyala;
University of British Columbia
Background
Advances in visualization software have allowed for stereoscopic 3D anatomy viewing using virtual reality devices, such as the HTC Vive and Oculus Rift. However, the practical use and uptake of such tools is blunted by financial and portability concerns. With increasing interest in using VR learning objects, lightweight, low-cost VR options such as smartphones equipped with Google Cardboard present strong alternatives to larger devices, especially in typical learning environments. In contrast to polygonal models typically used in anatomy model viewers, volume rendering allows the user to slice the model in order to visualize internal anatomy, and segment 3D anatomical structures into their constituent 2D planes. Previous research has indicated that viewing anatomy in stereoscopic 3D can provide for enhanced learning.
Educational Goals / Teaching Points
To introduce a low-cost VR solution utilizing stereoscopic 3D medical imaging volumes to enhance imaging anatomy learning.
Key Anatomic/Physiologic Issues and Imaging Findings/Techniques
Using the open-source ImageJ application, CT/MRI DICOM files were converted into a format that could be imported into a volume renderer implementation in the Unity game engine, which itself is free for educational/commercial use. The Google Cardboard mixed-reality plugin was used to enable VR, and the resulting Unity app was then exported to Android phones. A Google Cardboard headset with a built-in headset button is required to make use of the app. The application developed by the authors allows clinical imaging data (CT, MRI) to be viewed in VR space as 3D volumes, creating self-contained learning objects. The user can interact with the sliders using a button located at the top of the headset, rotating and slicing the model freely in x/y/z axes. One of the limitations of using Google Cardboard-based VR is that there is no capability for positional tracking for authentic translational movement around the model. Additionally, in order to enable import into Unity, the resolution of the initial clinical imaging files must be reduced, which produces a slightly lower fidelity volume render compared to desktop applications. However, the developed process does have the advantage of high framerates and quick loading times.
Conclusion
Google Cardboard mobile VR offers an inexpensive, portable, and relatively high-quality option for a stereoscopic 3D experience that is easily adapted for use in anatomy education. The author-developed Unity application brings volume rendering to mobile devices in a VR environment, expanding the range of educational possibilities. Combined with the Cardboard VR solution, the application takes advantage of the ubiquity of smartphone devices, helping with diffusion and uptake of VR anatomy education.
