Diagnostic Performance of a Next-Generation Virtual/Augmented Reality Headset: A Pilot Study of Diverticulitis on CT

Area of Science:

• Radiology and Medical Imaging
• Virtual Reality (VR) and Augmented Reality (AR) in Healthcare
• Diagnostic Performance Evaluation

Background:

• Next-generation VR/AR headsets offer potential for clinical interpretation of radiologic images.
• Validation is crucial for high-resolution, low-luminance diagnostic tasks like identifying diverticulitis.
• Desktop systems are currently the approved standard for radiologic image interpretation.

Purpose of the Study:

• To compare the diagnostic performance of radiologists in detecting diverticulitis on CT scans using a VR/AR headset versus a desktop system.
• To survey radiologists' user experience, ease of use, and device preference.

Main Methods:

• A pilot study retrospectively analyzed 110 patient CT scans (55 with diverticulitis, 55 controls).
• Six radiologists interpreted 100 unknown scans each on both a VR/AR headset (Apple Vision Pro) and a desktop.
• Diagnostic performance was assessed using ROC curves; time per case and user experience were surveyed.

Main Results:

• Area Under the Curve (AUC) for diverticulitis detection was similar: 0.93 (VR/AR headset) vs. 0.94 (desktop) (p=0.40).
• Time per case was significantly longer with the VR/AR headset (median 57 seconds) compared to the desktop (median 31 seconds) (p<0.001).
• Radiologist feedback on ease of use, experience, and preference varied, with mixed reviews for the VR/AR headset.

Conclusions:

• Next-generation VR/AR headsets demonstrate comparable diagnostic performance to desktop systems for identifying diverticulitis on CT.
• VR/AR headsets require further optimization to reduce interpretation time per case.
• User experience and preferences for VR/AR headsets in radiology require further investigation.