A novel method for separating the components of the clinical electroretinogram

Abstract

Ganzfeld electroretinograms (ERG) are summated responses of various layers of the retina to a flash of light. Pharmacological and non-invasive techniques are used in research labs to separate out the components arising from these layers. These methods, however, cannot be applied to ERG obtained in clinics for diagnostic purposes. Here, we outline the applicability of wavelet techniques to separate a-wave, b-wave and at least one oscillatory potential (OP) from the ERG obtained under scotopic conditions. Using white flashes, 62 ERG waves were analysed using Daubechies D4 wavelets. Using L ¹-norm as the cost function, the wavelet coefficients that carry the most information were identified. Using histograms of the identified coefficients and their scalograms, we extracted OP1 and a-wave, and the remainder could be approximated to be the b-wave. Improvement based on Breton's and Pepperberg's models for the separation of a-wave is also proposed.

Acknowledgements

This work was supported in part by a grant from the University of Missouri Research Board to VL. At the time this work was done, VL and LSV were affiliated with the University of Missouri in Saint Louis, MO.

Notes

†If the wavelet was scaled by N, then the level, j, is given by N = 2 ^j .