Purchase this article with an account.
Mathieu Gauvin, Jean-Marc Lina, Julie Racine, Pierre Lachapelle; Modeling A Progressive Retinal Degeneration With The Photopic Hill: A Time And Frequency Approach. Invest. Ophthalmol. Vis. Sci. 2011;52(14):694.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In response to a progressively brighter stimuli, the amplitude of the human photopic b-wave steadily increases to reach a maximum value following which a gradual decrease is noted despite even brighter flashes, a phenomenon known as the photopic hill (PH). We examined if the inverted PH (IPH: from Vmax to ERG to dimmest flash) could be used to model the progressive functional loss that characterizes several retinal degenerative processes.
This study was conducted on photopic ERGs from normal subjects [n=85; flash stimuli: -2.62 to 0.64 log cd.sec.m-2 in 17 steps of ~ -0.2 log-unit; background: 30cd.m-2] and from patients [n=55; flash intensity: 0.64 log cd.sec.m-2; background: 30cd.m-2; minimum of 3 visits; time interval: 5 to 25 years]. Descriptors derived from the discrete wavelet transform (DWT) were used to quantify the ERGs. Normal luminance-response (L-R) graphs were used to correlate the functional lost that time would produce in the presence of retinal degeneration. The cross-covariance was used to match normal and pathological ERG waveforms.
In normal, the IPH gradually decreased from 131.42±31.27µV (Vmax) to 0.71±0.12 µV (dimmest flash used) in two distinct pseudo-asymptotical steps: From 0.64 to -1.0 log cd.sec.m-2 [R²=0.9741, slope: -33.20±4.38µV.s per decrement; 9 steps total] and from -1.0 to -2.62 [R²=0.8996, slope: -1.0324±0.18µV.s per decrement; 9 steps total]. Of interest, a significantly higher coefficient of variation (CV=77%) was observed at the asymptotes intersection suggesting a light-sensitive threshold property of the retina. Pathological ERGs could always be fitted to this model. However, the intensity decrement equivalence between two consecutive ERG measures varied significantly between patients and did not appear to be time dependent. The latter was most obvious in conditions where the two eyes were affected differently at the onset. In patients where more than two ERG measures were obtained, our model was able to predict the time of occurrence of the next ERG value with a 98% accuracy.
The two distinct pseudo-asymptotical steps that describe the IPH were never reported. According to the preliminary results the IPH could be used to model the progression of some degenerative retinopathies, a feature that could possibly help in the staging and prognosis of these patients. Supported by FFB (USA).
This PDF is available to Subscribers Only