2017/05/22 by Stephan Wyder, Philippe C. Cattin, Wyder, Stephan +1 · 1 citation
Computer Science · Medicine · #FOS: Computer and information sciences #Gaze Tracking and Assistive Technology #Glaucoma and retinal disorders #Human-Computer Interaction (cs.HC) #Retinal and Optic Conditions
paper · pdf · doi:10.48550/arxiv.1705.07589
openalex publication_date 2017/05/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Purpose. We present a new method to evaluate the accuracy of an eye tracker\nbased eye localization system. Measuring the accuracy of an eye tracker's\nprimary intention, the estimated point of gaze, is usually done with volunteers\nand a set of fixation points used as ground truth. However, verifying the\naccuracy of the location estimate of a volunteer's eye center in 3D space is\nnot easily possible. This is because the eye center is an intangible point\nhidden by the iris. Methods. We evaluate the eye location accuracy by using an\neye phantom instead of eyes of volunteers. For this, we developed a testing\nstage with a realistic artificial eye and a corresponding kinematic model,\nwhich we trained with \μCT data. This enables us to precisely evaluate the\neye location estimate of an eye tracker. Results. We show that the proposed\ntesting stage with the corresponding kinematic model is suitable for such a\nvalidation. Further, we evaluate a particular eye tracker based navigation\nsystem and show that this system is able to successfully determine the eye\ncenter with sub-millimeter accuracy. Conclusions. We show the suitability of\nthe evaluated eye tracker for eye interventions, using the proposed testing\nstage and the corresponding kinematic model. The results further enable\nspecific enhancement of the navigation system to potentially get even better\nresults.\n