Image analysis tools for assessment of atrophic macular diseases

Abstract The explosion of new retinal imaging technologies and the exponential growth in the size and complexity of retinal imaging datasets has fueled a demand for automated retinal image analysis tools. One disease where image analysis tools are of particular importance is age-related macular degeneration (AMD). Image analysis approaches to identify and quantify atrophy associated with AMD have evolved from semiautomatic approaches requiring extensive manual input to more automated approaches using prespecified or custom-crafted features, to fully automated artificial intelligence (AI) deep learning approaches where image features are learned automatically. In addition, similar approaches appear to be applicable for the automated identification and quantification of risk factors in early/intermediate AMD eyes which may be associated with risk for progression to late AMD. This may allow the design of novel therapeutic trials aimed at intervention early in the disease process.