Physiological and noise components of the PERG intrinsic variability

Abstract

Purpose: To isolate and quantify components of the Steady-state Pattern Electroretinogram (SS-PERG) that are not explained by noise, in order to gain new physiological information and control the sources of intra-test variability.

Methods: Using a new device (Miami-PERG) SS-PERGs were recorded in 10 normal subjects (mean age 36 ±9 years) in response to horizontal black-white gratings generated on a LED tablet (14 x 14 cm, 800 cd/m² mean luminance, bar size 0.67 deg at 30 cm viewing distance) reversing 15.63 times/s. Signals were simultaneously recorded from both eyes with skin electrodes over the lower eyelids (reference ipsilateral temples). Signals were continuously acquired (1-300 Hz bandpass) over 2.5 minutes and averaged in 16 successive samples of 64 epochs each (1024 epochs for grand-average SS-PERG). For the SS-PERG waveform, as well as for that of subsamples, amplitudes/ phases of the response component at the reversal rate were evaluated by means of Fourier analysis. Corresponding noise measurements were obtained using the +/-reference method.

Results: The mean SS-PERG amplitude of subjects was 1158 ±191 nV, the mean phase was 66 ±10 deg (estimated latency 52 ±2 ms) and the mean noise was 59 ± 26 nV. The amplitude of subsamples displayed progressive decline of 222 ±166 nV (P<0.001) over 2.5 minutes acquisition period (adaptation) while the phase remained stable. On average, adaptation accounted for 29 ± 16% of the total amplitude variance of subsamples. The noise accounted for 51 ± 23% of the total amplitude variance of subsamples; the non-adaptive, residual amplitude variance (fluctuation) accounted for 15±23%.

Conclusions: The SS-PERG has gained interest in the clinical application because it provides advantages in terms of speed, robustness and sensitivity to disease compared to the standard Transient PERG. This new protocol further improves the signal-to-noise ratio and allows evaluation of three components of the response’s intrinsic variability, 1) noise, 2) adaptation, 3) fluctuation. Adaptation and fluctuation of the SS-PERG signal contribute to a relevant portion of the SS-PERG intrinsic variability and may have physiological significance in clinical applications of PERG independently of the standard measurements of amplitude and phase.