MERIT poster finds strong agreement on AMD lesion detection methods

By AI, Created 10:59 AM UTC, May 20, 2026, /AGP/ – MERIT will present ARVO 2026 data showing strong agreement between manual and semi-automated OCT methods for identifying iRORA, a high-risk precursor to geographic atrophy in age-related macular degeneration. The findings suggest machine-learning–assisted analysis could help standardize lesion detection in dry AMD research and trials.

Why it matters: - iRORA is a high-risk precursor to geographic atrophy in age-related macular degeneration, so better detection could improve how researchers track disease progression. - More reliable lesion identification may also help assess emerging therapies in dry AMD. - A semi-automated method could speed analysis and reduce variability compared with manual grading.

What happened: - MERIT will present a poster at the ARVO 2026 Annual Meeting in Denver on May 6, 2026, from 10:15 AM to 12:00 PM. - The poster is titled “iRORA in intermediate and advanced atrophic AMD: detection and assessment by 2 OCT-based methods.” - Investigators compared manual grading using en face ORC images with a machine-learning–assisted Layer Loss Analysis method. - The analysis covered 37 eyes with intermediate AMD or geographic atrophy. - The two approaches showed 88.6% agreement by lesion.

The details: - The Layer Loss Analysis method co-localizes ellipsoid zone and RPE loss. - The study found that disagreements were mostly limited to borderline size thresholds. - Authors on the poster include Ronald P. Danis, Carl Regillo, Alexander J. Shaer, Li Fan, Jonathan Loi, Yijun Huang, and David Bingaman. - The presentation will take place at the Colorado Convention Center. - MERIT described itself as a global provider of clinical and preclinical trial endpoint and technology services.

Between the lines: - The results suggest LLA may offer a practical semi-automated option for identifying iRORA in OCT imaging workflows. - Ronald P. Danis said the method should be evaluated in a larger longitudinal data set. - That caution matters because the current analysis was post hoc and small. - The data point toward better consistency, but not yet definitive validation for broad clinical use.

What’s next: - MERIT and its collaborators plan to discuss the findings at ARVO 2026. - The next step is testing the method in a larger longitudinal cohort. - Further study will determine whether the agreement seen here holds over time and across more patients. - MERIT also operates offices in Madison, Shanghai, Toronto, and Sydney, and works across ophthalmology, respiratory, oncology, and cardiology.

The bottom line: - MERIT’s ARVO poster suggests a machine-learning–assisted OCT approach can closely match manual grading for identifying a key AMD lesion, with larger studies needed before broader adoption.