Topographic and Quantitative Correlation of Structure and Function Using Deep Learning in Subclinical Biomarkers of Intermediate Age-Related Macular Degeneration

AI Generated Summary*

Twenty eyes of 20 patients with intermediate age-related macular degeneration underwent OCT and four microperimetry runs, two each on photopic MP-3 and mesopic MAIA devices, using one 45-point grid. Deep learning measured drusen volume, hyperreflective foci and ellipsoid zone thickness; experts annotated subretinal drusenoid deposits and outer nuclear layer thickness. Stimuli were co-registered with OCT. Across 3,545 points, thinner ellipsoid zone and outer nuclear layer, plus greater hyperreflective foci and drusen volumes, each correlated with lower sensitivity (p<0.001); layer effects varied by eccentricity. Mean sensitivity was 26.25 dB on MP-3 versus 22.63 dB on MAIA. In univariate analysis, points with subretinal drusenoid deposits averaged 23 dB versus 26 dB without. The cohort was small and lacked longitudinal data.

*This summary was generated by AI and is published unedited. Oku does not alter these summaries. It may contain errors or omissions and is provided for general informational purposes only. Oku does not guarantee its accuracy, completeness, or reliability. For authoritative information, please refer to the original, peer-reviewed article.