Purchase this article with an account.
Ye Chen, Jessica Tang, Marc Sarossy; Statistical Decomposition of the Electroretinogram into its Components. Invest. Ophthalmol. Vis. Sci. 2014;55(13):3511.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In the early 20th Century, Piper, Granit and others decomposed the electroretinogram (ERG) into components using pharmacological techniques. In this study, we decompose the ERG into early, mid and late components using statistical curve fitting techniques in R.
This study tests the reliability of a statistical decomposition of the photopic full-field ERG into a combination of Gaussian curves. 39 subjects with ‘normal’ functional eyes were recruited. Each subject underwent 2 sessions of bilateral ERG tests at five different brightness levels. Raw data obtained from the tests was recorded in the Espion software. This was exported to the R statistical program. A parametric function consisting of a linear combination of three Gaussian curves was fitted to the data using non-linear least squares.
Rapid convergence and successful curve fitting was achieved for 446 out of 457 traces. Reconstructed waveforms showed good correlation with the original raw data for a- and b-waves (r=0.945 for a-wave and r=0.972 for b-wave). Bland Altman plots also showed good agreement between modelled waveforms and raw data.
Statistical decomposition of the electroretinogram into early, mid and late components is possible and reliable. The analysis of individual components may prove useful in the study of particular diseases.
This PDF is available to Subscribers Only