Abstract
Abstract: :
Purpose: To determine whether a slow flash multifocal ERG (SF-mfERG) paradigm can detect dysfunction in diabetics with mild or no retinopathy, and to examine the retinal topographies of abnormalities identified by this technique. Since successive responses overlap less than with standard (fast) mfERG stimulation, the SF-mfERG could have advantages when late response components are examined. Methods: We tested 11 controls, 5 diabetics without retinopathy (D), and 9 diabetics with mild retinopathy (D+R; predominantly microaneurysms). SF-mfERGs were recorded from one eye of each subject (dilated pupils; 10-300 Hz filtering; VERIS 4 system). 103 areas of the central 45 deg were stimulated by flashes (100 cd/m^2) separated by 3 dark frames (75 Hz frame rate). Initially N1, P1 and N2 implicit times (ITs), and N1, N1-P1 and N2 amplitudes were determined for the mean first-order kernels of 6 concentric groups (group 1=central fovea; group 6=periphery) in each eye. First-order N1, P1 and N2 ITs and the scalar product amplitudes (SPAs) of the first- and second-order kernels were then derived from each of the 103 locations and mapped. Abnormalities were defined as values ≷= 2 standard deviations from the control means and by t-test (p<0.05). Results: [1] In the D+R subjects, 3 response groups had abnormal N1 ITs and 4 groups had abnormal N2 ITs. Their amplitude abnormalities were: N1 in 4 groups; N1-P1 in 2 groups; N2 in 3 groups. [2] D subjects had IT abnormalities in groups 5 (N1) and 1 (P1). None of their groups were abnormal in amplitude. [3] Retinal maps of D+R subjects showed abnormal ITs occurred most frequently in the parafoveal and inferior retina. The topographies of their SPA abnormalities differed between the response kernels. [4] The IT abnormalities of D subjects also occurred most frequently in the parafoveal region and their SPA abnormalities also differed between kernels. [5] Ancillary recordings showed 6 of 9 D+R and 3 of 6 D eyes had a greater number of N2 IT abnormalities in SF-mfERGs than in standard mfERGs. Conclusion: The SF-mfERG appears to be sensitive to mild diabetic retinopathy. It can also identify abnormalities in diabetics without retinopathy. Implicit time was a more sensitive indicator of abnormality than amplitude, and ITs of N1 and N2 were more sensitive than P1. Abnormalities of diabetic eyes occurred most frequently in the parafovea. SF-mfERGs may be more sensitive to diabetic dysfunction and retinopathy than standard mfERGs. We are currently comparing these techniques.
Keywords: 395 electroretinography: clinical • 388 diabetic retinopathy • 387 diabetes