Analyzing Age-Related Macular Degeneration Progression in Patients with Geographic Atrophy Using Joint Autoencoders for Unsupervised Change Detection

Abstract

<jats:p>Age-Related Macular Degeneration (ARMD) is a progressive eye disease that slowly causes patients to go blind. For several years now, it has been an important research field to try to understand how the disease progresses and find effective medical treatments. Researchers have been mostly interested in studying the evolution of the lesions using different techniques ranging from manual annotation to mathematical models of the disease. However, artificial intelligence for ARMD image analysis has become one of the main research focuses to study the progression of the disease, as accurate manual annotation of its evolution has proved difficult using traditional methods even for experienced practicians. In this paper, we propose a deep learning architecture that can detect changes in the eye fundus images and assess the progression of the disease. Our method is based on joint autoencoders and is fully unsupervised. Our algorithm has been applied to pairs of images from different eye fundus images time series of 24 ARMD patients. Our method has been shown to be quite effective when compared with other methods from the literature, including non-neural network based algorithms that still are the current standard to follow the disease progression and change detection methods from other fields.</jats:p>

American Society of Retina Specialists Clinical Practice Guidelines on Multimodal Imaging for Retinal Disease

“…[50][51][52] However, a major limitation is that the current algorithms poorly predict the directionality of progression (eg, toward the fovea). 53 Further advancements in this area may lead to additional clinical tools to prognosticate vision-threatening GA progression.…”

Unsupervised Approaches for the Segmentation of Dry ARMD Lesions in Eye Fundus cSLO Images

“…Using deep learning, the authors of [ 26 ] applied a joint auto-encoder (initially applied on satellite images [ 27 ]) on the same dataset we are using in this paper (see Section 2 ), in order to perform automatic change detection in an unsupervised context learning. The algorithm outperformed the state-of-the-art; however, as this model aims to detect changes, it is not comparable with automatic segmentation algorithm.…”

Accurate Diagnosis of Diabetic Retinopathy and Glaucoma Using Retinal Fundus Images Based on Hybrid Features and Genetic Algorithm

“…The first direction is the binary computer-aided diagnosis (CAD) system that is based on retinal blood vessel segmentation [13]. The second direction is binary CAD systems dedicated to identifying age-related macular degeneration (AMD) disease [9]. The third direction is binary retinal CAD systems mainly devoted to segmenting the optic disc (OD) region [3].…”

“…Luckily, images of the retina can help in diagnosing several retinal diseases, including DR and glaucoma [9]. This process is usually carried out and interpreted manually, which is laborious and prone to error due to minute image details, as shown in Figure 1.…”