Major advances in diagnostic technologies are offering unprecedented insight into the condition of the retina and beyond ocular disease. Digital images providing millions of morphological datasets can fast and non-invasively be analyzed in a comprehensive manner using artificial intelligence (AI). Methods based on machine learning (ML) and particularly deep learning (DL) are able to identify, localize and quantify pathological features in almost every macular and retinal disease. Convolutional neural networks thereby mimic the path of the human brain for object recognition through learning of pathological features from training sets, supervised ML, or even extrapolation from patterns recognized independently, unsupervised ML. The methods of AI-based retinal analyses are diverse and differ widely in their applicability, interpretability and reliability in different datasets and diseases. Fully automated AI-based systems have recently been approved for screening of diabetic retinopathy (DR). The overall potential of ML/DL includes screening, diagnostic grading as well as guidance of therapy with automated detection of disease activity, recurrences, quantification of therapeutic effects and identification of relevant targets for novel therapeutic approaches. Prediction and prognostic conclusions further expand the potential benefit of AI in retina which will enable personalized health care as well as large scale management and will empower the ophthalmologist to provide high quality diagnosis/therapy and successfully deal with the complexity of 21st century ophthalmology.

Introduction

No field in ophthalmology has been scientifically and clinically blessed as much as retina in recent years. Retinal disease is given intensive and widespread attention with a common understanding that the condition of the retina is among the leading causes of severe vision loss and blindness on the global level. Age-related macular degeneration (AMD) currently affects 170 million people world-wide (Pennington and DeAngelis, 2016), while diabetic retinopathy (DR) is recognized as a world-wide epidemic. A third of an estimated 285 million people with diabetes have signs of DR and one third of them have vision-threatening DR (Lee et al., 2015). Furthermore, the numbers are increasing: it is anticipated that 288 million people will have AMD by 2040 and the number with DR will triple by 2050. On the other hand, therapeutic improvements in retina count among the major break-throughs in modern medicine. The introduction of intravitreal vascular endothelial growth factor (VEGF) inhibition in 2006 hugely reduced legal blindness rates and achieved considerable improvements in vision in neovascular AMD and diabetic macular edema (DME) (Varma et al., 2015). However, the success story of anti-VEGF therapy in clinical studies comes with the bitter pill of largely inferior outcomes in the real-world setting (Mehta et al., 2018). This is mainly because of delays in identifying disease onset and progressive course, as well as the unpredictability of recurrence, which together derail long-term management, particularly in neovascular AMD, the most aggressive entity. Moreover, numerous phase II/III clinical trials in the most prevalent atrophic AMD type, which leads to irreversible loss of the central retina, have been disappointing. Even the inhibition of complement factors, believed to act as major drivers of geographic atrophy (GA), have failed to halt disease progression and vision loss, leaving the question of valid therapeutic targets and relevant biomarkers unanswered (Boyer et al., 2017). Hence, retinologists are struggling with long-term visual decline in large patient populations, health care providers face disproportionate budget drains and researchers are disheartened by failed trials.