Automatic Detection of Preferred Retinal Locus (PRL) for Low Vision Rehabilitation using Oculometrics Analysis

Low vision rehabilitation is an important step towards individuals with Age-related macular degeneration (AMD) regaining useful functional vision and quality of life. One of the key steps in low vision rehabilitation is the determination of the preferred retinal locus (PRL). Currently, localization of PRL is performed manually though the guidance of a low vision therapist and the process is highly time-consuming, labour-intensive and subjective. In this paper, we present an automated system to objectively and accurately locate an individual’s PRL with the aid of gaze tracking technology. We also propose a graph-based method for false fixation detection to improve the system robustness. We validated the reliability of the system by using eye tracking technology to simulate central vision loss on people with normal vision. Experimental results show the potential use of eye tracking in detecting PRL for low vision rehabilitation.