Clinical Validation of Virtual Reality–Based Oculography for Quantitative Ocular Alignment Assessment in Strabismus

Abstract Purpose To evaluate the clinical validity of a virtual reality (VR)-based eye-tracking system for quantitative ocular alignment assessment by comparing device-derived measurements with conventional clinical measurements in strabismus. Methods We retrospectively reviewed medical records of consecutive patients undergoing strabismus evaluation. We assessed ocular alignment using conventional prism cover testing and a VR-based oculography system (neos™, machineMD, Bern, Switzerland). We compared total deviation measured by alternate prism cover testing (APCT) and manifest deviation measured by simultaneous prism cover testing (SPCT) with device-derived measurements. We assessed associations using Spearman correlation and agreement using Bland–Altman analysis. Subgroup analyses were performed by horizontal deviation magnitude (<15 vs. ≥15 prism diopters [PD]). Results Forty-six patients (mean age, 37.61 ± 20.41 years; 25 [54.3%] female) were included. Device-derived measurements correlated significantly with conventional measurements for horizontal total deviation (Spearman ρ = 0.780, P < 0.001), vertical total deviation (ρ = 0.728, P < 0.001), horizontal tropia (ρ = 0.783, P < 0.001), and vertical tropia (ρ = 0.424, P = 0.003). Bland–Altman analysis showed closer agreement for horizontal than vertical measurements. Measurement differences did not differ significantly between the <15 PD and ≥15 PD groups, although greater variability was observed in the ≥15 PD group. Conclusions VR-based oculography demonstrated significant associations with conventional measurements, with better agreement for horizontal than vertical deviations. Greater variability with larger horizontal deviations suggests cautious interpretation in large-angle strabismus.