Automated Measurement of Strabismus Angle Using a Commercial Virtual Reality Headset.

Abstract

Current clinical measurements of strabismus angles (SAs) are manual and require a skilled examiner and active cooperation by patients. This results in high interexaminer variability, and clinical use is limited by the availability of trained examiners. An objective and automated procedure, independent of the examiner, would be useful. This single-center, prospective, diagnostic feasibility study compared the vertical and horizontal SA of patients, as measured with a commercially available virtual reality headset (VRH) and custom software with the gold standard measurements performed manually with the Harms tangent screen (HW) and the alternate prism cover test (ACT). We implemented the ACT by showing the patient a fixation target on each eye alternatively with the VRH while recording the eye position of the patient. We then processed the data with custom written software to calculate the SA of the patients. These measurements were then compared to the SA measured with HW and the clinical ACT. Thirty-three patients took part in our study. We found good correlation between the VRH method and the HW as well as the clinical ACT. Best correlation was found for horizontal SA in the primary position, with the vertical SA in the primary position also correlating well. Peripheral gaze resulted in a slightly lower correlation due to the overestimation of horizontal SA and underestimation of vertical SA; cyclorotation was not measured with the VRH. Overall, VRH, HW, and clinical ACT correlated similarly well as the published interexaminer correlation for ACT. The automated measurement of strabismus with a VRH is feasible, easily applicable, fast, accurate, and can be run on consumer hardware that is affordable and increasingly available.