Artificial intelligence based assessment of treatment response in wet age related macular degeneration using paired OCT angiography

Artificial intelligence (AI) has shown promise in retinal imaging, yet its application to longitudinal optical coherence tomography angiography (OCTA) remains limited. This study developed and evaluated an AI model for classifying treatment response in neovascular age-related macular degeneration (nAMD) using paired OCTA images acquired before and after anti-VEGF therapy. In this retrospective cohort study, paired OCTA en-face images and corresponding OCT B-scans were collected for each treatment course. OCTA image pairs were manually segmented and aligned for AI input, while ground-truth labels (Improved, Unchanged, Worsened) were determined based on structural OCT findings and clinical visual acuity outcomes. After quality exclusion, 1033 OCTA pairs were included and divided into training, validation, and independent testing subsets. Two experienced retina specialists graded all OCTA pairs for comparison. On the test set, the AI model achieved an overall accuracy of 82.08%, with class-specific accuracies of 74.29% (worsened), 81.48% (unchanged), and 88.64% (improved). In contrast, overall human grading accuracy was 61.40%. Human graders were significantly more likely to misclassify treatment response than the AI model for all groups (odds ratio = 2.88; 95% CI 1.68–4.92; p < 0.0001). These findings demonstrate that AI-based paired OCTA analysis can provide a more accurate and objective assessment of treatment response in nAMD.