Wearable Augmented Reality for Nystagmus Examination in Patients With Vertigo: A Randomized Crossover Usability Study
Introduction:
Vertigo, a common symptom with various underlying causes, often stems from benign vestibular dysfunction but can also be central, such as in stroke cases. Nystagmus analysis is crucial for differentiating these disorders, yet conventional video-oculography (VOG) demands specialized equipment and expertise, limiting accessibility. This study introduces a wearable augmented reality (AR) system for standardized oculomotor testing with real-time eye tracking, aiming to address these challenges.
Methodology:
The study evaluated the usability, accuracy, and tolerability of the wearable AR-based nystagmus system in a hospital setting. It integrated hardware and software for oculomotor testing, real-time eye tracking, and automated data processing. Nine patients with vertigo were enrolled, and eight completed both AR and VOG examinations in a randomized crossover design with a 30-minute washout period. Discomfort was assessed using a visual analog scale (VAS).
Results:
The AR system demonstrated diagnostic consistency comparable to VOG, particularly in ruling out central abnormalities. Agreement rates between AR and VOG ranged from 62.5% to 87.5%, with an overall diagnostic accuracy of 77.1%. Sensitivity and specificity were 81.8% and 75.7%, respectively. VAS scores did not differ significantly between AR and VOG, indicating tolerability.
Discussion:
This study showcases the feasibility of wearable AR glasses for nystagmus examination, offering portable and automated assessment. Results align with recent AR-based studies, highlighting its potential for point-of-care testing. However, limitations include small sample size and the need for improved processing and multicenter validation.
Controversy & Comment Hooks:
The study's small sample size and wide confidence intervals for sensitivity and specificity raise questions about precision. The moderate positive predictive value (PPV) underscores the need for enhanced processing. Additionally, the study's reliance on percent agreement for diagnostic concordance, rather than Cohen's κ or multiple raters, limits its rigor. These points invite discussion on the study's validity and the potential for future improvements.
