VR Based Strabismus Diagnosis using Image Processing

Abstract

Strabismus is an eye movement disorder that the eyes do not properly align with each other when looking at an object. Currently, the ophthalmic prism with two nonparallel planes is used to diagnose the strabismus angle. When the proper prism is placed over the misaligned eye, there is no movement of the eyes when the cover is switched back and forth between the eyes. The strabismus angle is then identified after checking the parameter of the prism. However, the whole process is time-consuming and the results depend on the doctors’ experience. In this study, an automated strabismus diagnosis technique using a VR device is developed. A VR display system is built to simulate the normal strabismus diagnosis steps, in which screens are changed alternately between on and off stages. Patients watch specially-designed 3D scenes, and the eye motions are recorded by two IR cameras. An image-processing based pupil tracking technique is then applied to track the eye movement. After recording the eye position, two strategies for strabismus angle estimation, stepwise approximation and direct measurement, are implemented. The stepwise approximation measures the angle with the feedback calibration process, while the direct measurement converts the eye displacement to the strabismus angle after considering the eyeball diameter. With the proposed process, the strabismus angle is accurately identified. To demonstrate the performance of these techniques, experiments are carried out on various persons, including strabismus patients. The results are compared to the results of doctor’s diagnosis, which clearly indicates that these techniques could identify the strabismus angle with high accuracy and efficiency.