How imaging leaders can benefit from early AI adoption
An introduction to VR and AR
VR and AR are both image display techniques that present holograms—three-dimensional, computer-generated images—to the user via a wearable device such as a headset or glasses.
- VR uses a fully enclosed headset to replace the outside environment with a complete virtual world. This allows the user to interact with and manipulate the projected holograms and the virtual world in an immersive experience.
- AR relies on glasses-like devices to overlay computer-generated holograms on top of the user's current environment. This allows the user to interact with the computer-generated projections while simultaneously continuing to experience the world around them.
While these technologies have been in development for some time, they've recently grown in popularity in the gaming and entertainment world. This has led to lowered prices and improved computing power, resulting in the exploration of potential use cases in medicine.
3 potential uses in radiology
1. Education: Radiologist education is perhaps the most obvious application for VR in radiology. It's also likely to be the first broad-based use. VR-based education could allow radiologists or residents to view a 3-D image in real time, manipulate the image to get different perspectives and views, and even interact with the image. For example, residents could use a VR headset to view the 3-D anatomy of an image and walk around the image to view different angles. In many ways, the use of VR in education is similar to the growing use of 3-D printing for educational purposes.
2. Diagnostic image reading: VR has the opportunity to change both how radiologists view images and where they can view these images. As VR images continue to improve, 3-D rendered images could become the new "normal," allowing radiologists to manipulate the hologram during the reading process to access additional information. Further, VR headsets will provide additional flexibility to radiologists by replacing the need for a workstation; radiologists will be able view images from home or remote locations.
3. Image-guided interventional procedures: While the fully-immersive environment of VR may not be conducive to interventional procedures and surgeries, AR has the potential to dramatically change how these procedures are conducted. For example, an AR application could project a 3-D hologram generated from an MRI projected onto a patient's body. This reduces, or potentially eliminates, the need for the physician to look up from the procedure table at prior imaging. In preliminary studies of such techniques, AR has been shown to maintain the accuracy of the guidance, while reducing both procedure time and exposure to radiation.
Current challenges
While there is great potential for the use of AR and VR in imaging, it is important to recognize that there are still hurdles that must be cleared before such applications hit the market and are widely used in medicine.
For one, the quality of images displayed using AR and VR devices is lower than the quality that radiologists are accustomed to working with. Additionally, current headset devices may weigh up to a pound, which can lead to discomfort for physicians if used over an extended period of time. And motion sickness remains a challenge, especially in VR, which means that some physicians may never feel comfortable using VR as part of day-to-day practice. Finally, VR and AR applications in imaging may struggle to overcome many of the same regulatory hurdles as AI. VR viewing of medical images and use of AR to project images or support interventional procedures are not currently FDA approved, and this may be the final hurdle for vendors before they begin to flood the medical market with products.
This article is based in part on conversations and presentations from the 2017 Economics of Diagnostic Imaging Conference held in Arlington, Virginia.