A Video-Oculography Based Telemedicine System for Automated Nystagmus Identification

Abstract

This paper presents the research and results for an automated jerk-type nystagmus identification system that makes use of an efficient, low cost video-oculography (VOG) device, designed for telemedicine applications. The pupil position is estimated by a hybrid tracking algorithm from the captured VOG images. It is also shown that wavelet analysis with an appropriate mother wavelet, coupled with well-defined geometric constraints can provide a reliable and robust nystagmus identification algorithm. Some original research regarding robust analysis for signals with mixed content is also presented.