Purchase this article with an account.
J. G. Robson; Quantifying the Waves in Electroretinograms of Low Amplitude. Invest. Ophthalmol. Vis. Sci. 2007;48(13):1517.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To devise objective methods of determining the times and amplitudes of the principal waves in low amplitude electroretinograms (ERGs) having poor signal to noise ratio (S/N) e.g. pattern ERGs (PERGs) and focal macular ERGs (FMERGs) obtained with low frequencies of stimulation.
1) Examination of the energy spectra of transient PERGs and FMERGs conventionally recorded with a high-frequency bandwidth limit of 100 Hz or more, indicates that the principal waves are related to components with frequencies less than about 50 Hz. Thus the S/N of such records can be improved by removing all components with frequencies of 50 Hz and above. This can be done by first obtaining a representation of the recorded ERG as a Fourier series and subsequently regenerating a filtered version of this ERG by summing only those members of the series whose frequency is lower than 50Hz. The amplitudes and peak times of the various waves are then assigned the values of the (dominant) maxima and minima within defined time windows. 2) For low amplitude ERGs that are expected to have only 3 principal waves (PERGs and FMERGs) the amplitude and time of the troughs and peak can be ascertained by fitting a suitable model function that has the appropriate triphasic form and whose parameters can be varied to provide a good fit to the whole waveform. One such model is provided by summing three functions of the kind V = A (t/τ)ne-t/τ (a variant of the probability density function of a gamma distribution) where the three parameters A, n and τ for each function can be independently adjusted to give the best fit.
Application of these operations to many published records of PERGs and MFERGs, as well as to other recordings obtained from various sources, indicates that acceptable values of the amplitudes and peak times of the principal waves can normally be obtained by either of these methods though only the second method guarantees that there will never be any ambiguity in the assignments.
Estimation of the times and amplitudes of the principal waves of ERGs with poor signal to noise ratio can be satisfactorily performed by the application of purely mechanical methods that can be precisely specified. These methods provide an objective way to implement the instruction "the peak [of the PERG] should be designated where it would appear on a smoothed or idealised waveform" that is given in the Standard for Basic Pattern Electroretinography 2006) issued by ISCEV (International Society for Clinical Electrophysiology).
This PDF is available to Subscribers Only