Impact of notch filter use on waveforms of First- and Second-Order-Kernel responses from multifocal ERGs

The filter settings of the amplifier section of ERG recording systems have large impact on the waveforms of the recorded responses. In this study, the effects of a 50 Hz notch filter were assessed for both First-Order-Kernel- (FOK) and Second-Order-Kernel-responses (SOK 1st slice) from multifocal ERGs recorded with a VERIS III system. Amplitude and phase responses of the amplifier section were recorded for typical filter settings with the notch filter active and inactive. Multifocal ERGs (MERGs) from a group of 11 normal subjects were recorded for both amplifier settings, the waveforms resulted from those recordings were compared in the time and frequency domain. To verify the results, the recordings without the notch filter were digitally filtered with a simulated notch filter and compared to the responses recorded with an active analog notch filter. The line filter has the biggest attenuation at 50 Hz with an additional phase jump of 180 degrees. The FOK responses of MERGs are assembled by frequencies below 65 Hz with main spectral components between 19 and 47 Hz. The 1st slice of the SOK consists of frequencies up to 100 Hz with main components between 19 and 84 Hz. Thus, if FOK recordings are to be analyzed, the notch filter of the amplifier can be used in order to cope with noise problems caused by the line frequency of 50 Hz. However, one must be aware that the attenuation of higher frequency components will lead to changes of the waveforms. For SOK analysis, main spectral components are attenuated and/or shifted in phase, which leads to completely different waveforms and severe distortion of the recording results.