Validation of virtual reality system based on eye-tracking technologies to support clinical assessment of glaucoma

Aim: The aim was to develop and implement a virtual reality tool based on eye-tracking technologies that allow to evaluate the characteristics of the gaze patterns of glaucoma patients in order to have a better understanding of the limitations that these patients experience in their daily life. Setting: This study took place on the Ophthalmology department of Hospital Clinico San Carlos, Madrid, Spain. Methods: In total, 56 participants collaborated in the study. They were divided in two groups, a group composed of 33 glaucoma patients selected by the Ophthalmology department and a control group composed of 23 healthy individuals. Both groups completed two virtual tasks while their gaze was being monitored. The first task, defined as "static" consisted in two exercises based on the observation of images. The second task, defined as "dynamic," consisted in a virtual driving simulator. Number of fixations, fixations duration, saccades amplitude and velocity, fixations/saccades ratio, total execution time, and other specific metrics were measured. These are the total search time for the second exercise of the first task and the number of collisions for the dynamic task. In addition, the dispersion of fixations was also discussed. Results: For the two exercises of the static task, patients exhibited significative differences in terms of number of fixations (p = 0.012 in free observation exercise), mean saccadic velocity (p = 0.023 and 0.017), fixations/saccades ratio (p = 0.035 and 0.04), and also the search and total execution times of the visual search exercise (p = 0.004 and 0.027, respectively). For the dynamic task, significative differences were found on average saccades amplitude (p = 0.02), average saccades velocity (p = 0.03), and the number of collisions (p = 0.02). Conclusion: The results show that eye-tracking technologies can be used as a tool for evaluating the gaze patterns of glaucoma patients and differentiate them of healthy individuals. However, further studies with a larger cohort of participants and additional tasks are needed.