Artificial Intelligence Algorithms for Analysis of Geographic Atrophy

Artificial Intelligence Algorithms for Analysis of Geographic Atrophy

Age-related macular degeneration (AMD) is a leading cause of vision loss in population over 50 years age in the United States. The advanced stage of AMD can be wet (also known as neovascular AMD) or dry (also known as geographic atrophy, GA). There are currently approved treatments for wet AMD but no known treatments for GA. The natural history of GA, which presents as atrophic patches in the retina, has been well studied in many clinical trials. However, it is still unclear what initiates the development of GA and why some lesions grow faster than others. The BAM study has a unique dataset of images obtained for the clinical trial including retinal photographs and specialized imaging called autofluorescence and OCT scans. The autofluorescence imaging and OCT scans provide insights into the retina beyond what the ophthalmologist can see and provide a means to fill the knowledge gap. Understanding development and growth of GA is key to developing targeted treatments. The purpose of this study is to use artificial intelligence (AI) algorithms to identify additional risk factors beyond human perception in the natural history of geographic atrophy. AI has to date mostly been used for automated segmentation of GA. We aim to take this further by developing AI models for identifying risk factors and morphological features for future GA development, predict rapid progressors, and possible relationship to visual function.

[{ "PostingID": 19810, "Title": "GSK-BAM114341", "Description": "A phase 2, multi-centre, randomised, double-masked, placebo-controlled, parallel-group study to investigate the safety, tolerability, efficacy, pharmacokinetics and pharmacodynamics of GSK933776 in adult patients with geographic atrophy (GA) secondary to age-related macular degeneration (AMD)" }]

Amitha Domalpally, Robert Slater, Rachel E. Linderman, Rohit Balaji, Jacob Bogost, Rick Voland, Jeong Pak, Barbara A. Blodi, Roomasa Channa, Donald Fong, Emily Y. Chew. Strong versus Weak Data Labeling for Artificial Intelligence Algorithms in the Measurement of Geographic Atrophy, Ophthalmology Science, Volume 4, Issue 5,
2024
DOI: https://doi.org/10.1016/j.xops.2024.100477.