Pupillometric recordings to detect glaucoma eyes

Objective. Glaucoma is the second cause of vision loss with early diagnosis having significantly better prognosis. We propose the use of hippus, the steady-state pupil oscillations, obtained from an eye-tracker for computerised detection of glaucoma. Approach. Pupillary data were recorded using a commercial eye-tracker device directly to the laptop. A total of 40 glaucoma patients and 30 age-matched controls were recruited for the study. The signals were de-noised, and the entropy of the steady-state oscillations was obtained for two light intensities, 34 and 100 cd m(-2). Main results. The results show that at 100 cd m(-2), there was significant difference (p < 0.05) between the sample entropy of the healthy eyes (0.55 +/- 0.017) and glaucoma eyes (0.7 +/- 0.034). The results at 34 cd m(-2) were also significantly different, though to a lesser extent. Significance. Entropy of the pupillary oscillations, or hippus, obtained using an eye-tracking device showed a significant difference between glaucoma and healthy eyes. The method used commercially available inexpensive hardware and thus has the potential for wide-scale deployment for computerized detection of glaucoma.