Estimation of eye condition using waveform shapes of pupil light responses to chromatic Stimuli

This paper examines the possibility of detecting 2 conditions which cause vision to deteriorate: Aged-Related Macular Degeneration (AMD), and the effects of aging on eyes using the features of PLR waveforms. These features were extracted using Fourier descriptors of PLR waveform shapes, weighted amplitudes of the waveforms, and a balanced combination of these two in the from of a weighted value. The Random Forest method was used for classification analysis to detect three types of PLR, such as in healthy eyes, in AMD-affected eyes, and in age-affected eyes. The optimized weight values were evaluated using a classification error rate. The results show that the error rates for healthy PLRs and AMD PLRs were low, but the error rates for PLRs of age-affected eyes stayed at a high level. Additionally, dissimilarities between the PLRs for blue light and red light at low intensities contributed to the performance of the classification technique.