Comparison of ERG signal-to-noise ratios in simultaneous recordings with skin electrodes and contact lens electrodes.

Purpose: To compare signal-to-noise levels in ERG recordings obtained with contact lens electrodes and adhesive skin electrodes.

Methods: 23 subjects were studied. Full-field ERGs were recorded according to ISCEV standards simultaneously with ERG-jet corneal contact lens electrodes and LKC Technologies Sensor Strip adhesive skin electrodes. B-wave amplitude or peak-to-peak amplitude was used as a measure of signal strength. Noise was estimated using the " ± averaging method." Comparisons between signal strength, absolute noise levels, and signal-to-noise ratios between contact lens and skin electrodes were performed by linear regression.

Results: Comparisons of signal strength for LA 3, 30-Hz, DA 0.01, and DA 3 responses, yielded regression coefficient ß values of 0.37, 0.39, 0.39, and 0.35, respectively. For the entire data set, the regression coefficient ß value was 0.36 (95% confidence limits 0.34 - 0.38). The grand average ERG noise for all ERG stimuli was 13.8 µV for contact lens electrodes and 13.0 µV for skin electrodes (not significant: p = 0.66 for paired t-test). For signal-to-noise ratios, regression ß coefficients for contact lens and adhesive skin electrodes for LA 3, 30-Hz, DA 0.01, and DA 3 stimuli were 0.25, 0.39, 0.50, and 0.36 respectively. The ß coefficient for the amalgamated data set was 0.33 (95% confidence limits 0.30- 0.36).

Conclusions: Overall ERG amplitudes obtained with skin electrodes were 1/3 those obtained with contact lens electrodes. Absolute noise levels were similar. Signal-to-noise levels with skin electrodes were 1/3 those seen with contact lens electrodes. Implications for signal-averaging in clinical applications are discussed.