Abstract
Purpose :
ERG responses to flash and step stimuli include high-frequency oscillations (OPs) that can be isolated with a bandpass filter of 75-300 Hz. Responses to pattern stimuli (pERG), generated with progressive-scan displays (e.g. CRT, LCD), do not typically contain OPs, regardless of the recording amplifier passband. Pattern ERG responses recorded with a synchronously-updated display do contain high-frequency oscillations well isolated with a bandpass filter of 50-300 Hz. This study evaluates the effect of progressive-scan displays on the recorded response waveform.
Methods :
pERG responses were recorded from six healthy eyes, using a custom synchronous LED-based display and a progressive-scan display (Diagnosys Espion with Pattern Stimulus Generator, refresh rate 100 Hz). Stimulus parameters of both displays were: check size = 10 degrees; mean ON luminance = 90 ph cd m-2; reversal rate = 4 RPS. Recording passband was 1-100 Hz for the progressive-scan display, and 1-300 Hz for the synchronous display. Response waveforms obtained with the synchronous display were used to simulate the response obtained with the progressive-scan display, by replicating the waveform ten times, shifting each copy in +1 ms increments so that the shifted reversal times spanned the range 0 < t < 10 ms, and then averaging the shifted copies, to simulate the ms-to-ms contribution to the recorded waveform during the 10 ms refresh period of the progressive-scan display. The simulated progressive-scan response waveforms were then compared to the actual progressive-scan response waveforms.
Results :
The high-frequency oscillations, which were prominent in the synchronous display waveforms, were nearly absent in the simulated progressive-scan waveforms. The simulated progressive-scan display waveforms matched those obtained with the actual progressive-scan display (0.89 < r2 < 0.98) more closely than did those obtained with the synchronous display (0.61 < r2 < 0.78). Filtering synchronous display waveforms with progressively narrower passbands until they matched the progressive-scan waveforms demonstrate that the progressive presentation of the pattern reversal to the retina over a 10 ms period has an effect equivalent to applying a 70 Hz low-pass filter.
Conclusions :
Progressive-scan displays with refresh rates of 100 Hz or less preclude the direct recording of high-frequency (significant energy above 70 Hz) response components in the pERG.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.