Abstract
Purpose: :
Identify artifacts and deconvolve the electrically evoked electroretinogram (eERG) in retinal implant wearers.
Methods: :
Four subjects implanted with an Argus II retinal prosthesis in the right eye (OD) participated in this study. Binocular eERGs were obtained, using Burian-Allen contact lens electrodes, by averaging up to 2750 epochs. Current levels ranged from below perceptual threshold to a maximum of ~50 µA/electrode. Signal-to-noise ratios of raw eERGs were increased off-line using wavelet transformation (WT). The eERG was expected to be measurable only in OD. Eye movements and pupil responses may contribute to the eERG, but will also evoke a response in the contralateral eye (OS), since they are centrally controlled. Therefore, pupil and eye movements were recorded with an eye tracker, averaging up to 30 responses.
Results: :
In three subjects eERGs could be recorded, which consisted of a negative peak (N1) followed by a positive peak (P1). In the fourth subject strong sinusoidal artifacts were present in the eERG that could not be sufficiently reduced with WT. In the other subjects eERGs could be obtained bilaterally. In one subject we performed eye tracking and eERG recordings before and after eye dilation with tropicamide (1%) and phenylephrin (2.5%). At 30 µA, the pupil dilated in both eyes at 0.6 s, followed by a bilateral constriction after 1.2 s, with smaller amplitudes in OD than in OS (-0.1, +0.3 vs. -0.2, +0.5 mm). Pharmacologic dilation abolished these pupil responses. Eye movements were small (0.1 mm or less). Before dilation, the eERG N1-P1 amplitude was 6 µV in both eyes. After dilation, the eERG amplitude was 2 µV in both eyes. To remove bilateral electrically evoked artifacts we subtracted the OS eERG from the OD eERG. No dilation drops were applied. Using this subtraction procedure we obtained reliable eERGs in three subjects. At current levels between perceptual threshold and maximum comfort level, eERG amplitudes were 2 - 5 µV, N1 latencies were 100 - 200 ms, and P1 latencies 300 - 400 ms. eERG amplitudes correlated significantly (F-test, P< 0.05, r2> 0.9) with stimulus level in 2 out of 3 subjects.
Conclusions: :
Bilateral artifacts, such as pupil responses, in the corneal eERG cannot be sufficiently reduced by using dilation drops. Even after dilation, a residual, but substantial electrical response persisted in the contralateral eye. This residual activity might reflect the neural component of the pupil reflex, while the myogenic component is blocked after dilation. Subtracting the contralateral eERG may yield the best approximation of the eERG.
Keywords: electroretinography: non-clinical • electrophysiology: non-clinical • retina