RESUMEN
PURPOSE: Retinitis pigmentosa (RP) is a group of degenerative retinal diseases characterized by mutations in genes affecting retinal pigment epithelium (RPE) function, as well as mutations directly involving photoreceptors. This paper aims to evaluate a nonlinear method to distinguish between the RP and normal eye based on the Electroretinogram (ERG) signal. METHOD: ERG signal was recorded from 28 eyes of patients with the RP and 32 normal eyes. The ERG signal consists of four different stimuli, including two dark-adapted and two light-adapted stimuli. The time-domain analysis includes the amplitude and implicit time to consider the robustness of the nonlinear method. A parabolic mapping method was performed, and two criteria (Theta angle and density) extracted from the parabola were compared for both groups. RESULTS: The results showed that a-wave's amplitude and implicit time significantly changed in the dark- and light-adapted stimuli. The amplitude of the b-wave showed significant changes in all stimuli. However, the implicit time of b-wave had a significant increase only in the dark-adapted 3.0 ERG. Both nonlinear criteria showed significant changes in the RP group for all the stimuli. The p-values of dark-adapted 3.0 (p = .0121), dark-adapted 10.0 (p = .0014), light-adapted 3.0 (p = .0119), and flicker 30 Hz (p = .0323) showed significant differences. Using the density criterion, the statistical test demonstrated a significant difference between the RP and healthy normal group in dark-adapted 3.0 (p = .0076), dark-adapted 10.0 (p = .0024), light-adapted 3.0 (p = .0021), and flicker 30 Hz (p = .0165). CONCLUSION: The proposed features have made it possible to distinguish between healthy and RP eyes. This method might be helpful in early diagnosis.