Abstract
Abstract: :
Purpose: The human pattern ERG (PERG) is supposed to be driven by nonlinearities of the local luminance response and by lateral interaction mechanisms. We here present a double frequency method to isolate a pure lateral interaction component (liPERG) which may be a more specific ganglion cell response than the standard PERG. Methods: 13 eyes of 13 visually normal subjects were examined. PERGs to vertical sinusoidal gratings of 3 different spatial frequencies (sf=0.056cpd, 0.22cpd, and 0.89cpd) were recorded monocularly with DTL electrodes. Neighboring vertical bars were modulated sinusoidally, but with different temporal frequencies (f1=8.3Hz and f2=6.7Hz). Thus the Fourier components at f1-f2=1.7Hz and f1+f2=15Hz indicate a nonlinear interaction between the vertical bars and were used to define the liPERGs. We analyzed the effect of spatial frequency on the liPERGs by applying a repeated measures ANOVA with post-hoc Scheffe test to the Fourier magnitudes at 1.7Hz and 15Hz. The significance values of the liPERGs (criterion p<1%) were quantified by the signal-to-noise ratio to the average magnitude at the 2 neighboring frequencies. Results: (1) Only 1 of 39 liPERGs at 1.7Hz was significant. (2) 34 of 39 liPERGs at 15Hz were significant. The magnitude (mean±SEM) and the number of significant responses (n) for the 3 different spatial frequencies were 0.610.09µV, n=10 (sf=0.056cpd), 0.720.11µV, n=11 (sf=0.22cpd), and 1.200.15µV, n=13 (sf=0.89cpd). (3) The statistical analysis showed a significant magnitude difference of the liPERGs at 15Hz between sf=0.89cpd and the two other spatial frequencies (ANOVA p<0.0001, Scheffe test p<0.0001). Conclusion: (1) Highly significant liPERGs were found for all subjects at 15Hz. (2) The liPERGs at 1.7Hz were difficult to isolate, probably due to noise contaminations by eye movements at low temporal frequencies. (3) The spatial frequency tuning of the liPERGs is consistent with earlier findings as significantly larger liPERGs were found for the finest grating compared to the two coarser gratings. (4) The large portion of significant responses at sf=0.056cpd indicates that lateral interaction mechanisms contribute to the human PERG even for low spatial frequencies. (5) The double-frequency method can be used to isolate a pure lateral interaction component in the human PERG. The liPERGs may be a helpful tool to study ganglion cell function, e.g., in glaucoma patients.
Keywords: 396 electroretinography: non-clinical • 395 electroretinography: clinical • 415 ganglion cells