Abstract
PURPOSE: To compare the abilities of different mfERG measurements to detect diabetic retinopathy. METHODS: Responses from 11 eyes of 11 patients with diabetic retinopathy and 20 eyes of 20 age-matched controls were recorded using VERIS 4 and standard conditions. First order mfERG scalar product amplitude (SPA), and P1 and N2 implicit times (ITs) were measured at the 103 stimulated retinal locations. Using a “stretching” method (Hood and Li, 1997), the amplitude and IT of the first positive peak were also derived. Abnormality was defined as Z-scores >= 2. Sensitivity of the 5 measures was evaluated by (1) counting the number of abnormal mfERGs and (2) determining correspondence with zones containing retinopathy. A zone was defined as the putative mfERG location plus its immediate neighbors based on fundus photographs. The Z-score of each zone was assigned by maximum Z-score. RESULTS: Amplitudes obtained by the stretching method and SPA detected few abnormalities, with little correspondence to retinopathy. From 1,133 mfERGs (103*11 eyes), the number of abnormalities detected by measuring IT were: 242 by the stretching method, 191 by P1 IT, and 130 by N2 IT. Evaluated by counting the most abnormalities in each eye, ITs determined by stretching were more than twice as effective as P1 and N2 ITs. The three IT measures showed similar sensitivity to severe retinopathy (detecting 3 or all 4 edematous areas), but in milder cases such as microaneurysms, the stretching method IT had the highest sensitivity (63%) while P1 and N2 ITs were ~40%. CONCLUSION: Among the measures of the mfERG we examined, implicit time obtained by the stretching method has the best sensitivity and spatial correspondence to diabetic retinal pathology.
Hood, D.C. & Li, J. (1997) A technique for measuring individual multifocal ERG records. In:Yager D, ed. Non-invasive assessment of the visual system, Trends in Optics and Photonics. Vol.11, Washington, DC: Optical Society of America; 280–2833+