Deep Learning-Based Classification of Inherited Retinal Diseases Using Fundus Autofluorescence

Alexandra Miere,Kawther Taibouni,Carlotta Pallone,Thomas Le Meur,Eric Petit,Vittorio Capuano,Oudy Semoun,Polina Astroz,Sylvain Berlemont,Donato Colantuono,Eric Souied,Yasmina Chenoune,Karen Bitton

doi:10.3390/jcm9103303

Abstract

Background. In recent years, deep learning has been increasingly applied to a vast array of ophthalmological diseases. Inherited retinal diseases (IRD) are rare genetic conditions with a distinctive phenotype on fundus autofluorescence imaging (FAF). Our purpose was to automatically classify different IRDs by means of FAF images using a deep learning algorithm. Methods. In this study, FAF images of patients with retinitis pigmentosa (RP), Best disease (BD), Stargardt disease (STGD), as well as a healthy comparable group were used to train a multilayer deep convolutional neural network (CNN) to differentiate FAF images between each type of IRD and normal FAF. The CNN was trained and validated with 389 FAF images. Established augmentation techniques were used. An Adam optimizer was used for training. For subsequent testing, the built classifiers were then tested with 94 untrained FAF images. Results. For the inherited retinal disease classifiers, global accuracy was 0.95. The precision-recall area under the curve (PRC-AUC) averaged 0.988 for BD, 0.999 for RP, 0.996 for STGD, and 0.989 for healthy controls. Conclusions. This study describes the use of a deep learning-based algorithm to automatically detect and classify inherited retinal disease in FAF. Hereby, the created classifiers showed excellent results. With further developments, this model may be a diagnostic tool and may give relevant information for future therapeutic approaches.

Highlights

Inherited retinal diseases (IRDs) encompass a large, clinically and genetically heterogeneous cluster of diseases that affect around 1 in 3000 people, with a total of more than 2 million people worldwide [1]
fundus autofluorescence (FAF) can capture the high-concentration intracellular A2E as hyperautofluorescent small lesions, corresponding to the flecks typically seen in Stargardt disease [12,13]
We describe a preliminary study to evaluate the use of deep learning for the automated classification of FAF images from a cohort of patients with Stargardt disease (STGD), Best disease (BD), and Retinitis Pigmentosa (RP)

Summary

Introduction

Inherited retinal diseases (IRDs) encompass a large, clinically and genetically heterogeneous cluster of diseases that affect around 1 in 3000 people, with a total of more than 2 million people worldwide [1]. IRDs are relatively rare, with an estimated prevalence of 1 in 16,500 to 1 in 21,000 for Best disease [16], 1 in 8000 to 1 in 10,000 for Stargardt disease [2,17], and 1 in 4000 for retinitis pigmentosa [18] Given their rarity, these conditions are often difficult to diagnose and patients can endure a long journey involving many ophthalmologists. Deep learning approaches for automated image analysis require large volumes of high-quality training data, which may be a challenging premise in a clinical setting These high volumes of data are even more difficult to obtain in the case of IRDs due to the rareness of these genetic conditions. We describe a preliminary study to evaluate the use of deep learning for the automated classification of FAF images from a cohort of patients with Stargardt disease (STGD), Best disease (BD), and Retinitis Pigmentosa (RP)

Datasets

Development of a Deep Learning Classifier

Findings

Discussion