As population aging has become a major demographic trend worldwide, patients suffering from eye diseases are expected to increase steeply. Early detection and appropriate treatment of eye diseases are significant in preventing vision loss and promoting living quality. Conventional diagnosis methods tremendously depend on physicians' professional experience and knowledge, which leads to a high misdiagnosis rate and a huge waste of medical data. Deep integration of ophthalmology and artificial intelligence (AI) can revolutionize disease diagnosis patterns and generate a significant clinical impact.

The rapid rise in AI technology requires physicians and computer scientists to have a good mutual understanding of the technology and the medical practice to enhance medical care in the near future.

AI applications

In recent years, AI techniques have shown to be an effective diagnostic tool to identify various diseases in healthcare. AI applications can make great contributions to support patients in remote areas by sharing expert knowledge and limited resources. While the accuracy of the models is incredibly promising, we need to remain prudent and sober when considering how to deploy these models in the real world.

Most studies regarding intelligent diagnosis of eye diseases focused on binary classification problems, whereas in clinical settings, visiting patients suffer from multicategorical retinal diseases. For instance, a model trained to detect AMD will fail to consider a patient with glaucoma as diseased because the model only can discriminate AMD from non-AMD. Choi and his colleagues applied DL to automatically detect multiple retinal diseases with fundus photographs. When only normal and DR fundus images were involved in the proposed DL model, the classification accuracy was 87.4%. However, the accuracy fell to 30.5% when all ten categories were included. It indicated that the model's accuracy declined while the number of diseases increased. To further enhance the applicability of AI in clinical practice, we should make more efforts to build intelligent systems that can detect different retinal diseases with high accuracy.

Additionally, in clinical practice, a single abnormality detected from one imaging technique cannot always guarantee the correct diagnosis of a specific retinal disease (e.g., DR or glaucoma). Multimodal clinical images, such as optical coherence tomography, angiography, visual field, and fundus images, should be integrated to build a generalized AI system for more reliable AI diagnosis.

Availability of data

The need for massive amounts of data remains the most fundamental problem. Although various data sets have been available, they only incorporate a small part of the diseases humans suffer from. Images with severe diseases or rare diseases are particularly insufficient. The population characteristics, various systematic disorders, and diverse disease phenotypes should be considered when selecting input data. Larger data sets from larger patient cohorts under different settings and conditions, such as diverse ethnicities and environments, are also needed in some automated diagnosis systems with impressive outcomes for further validation. 

Above all, by building interpretable systematic AI platforms using sufficient high-quality and multimodal data and advanced techniques, we can enhance the applicability of AI in clinical circumstances. Some days, we might make it possible to adopt intelligent systems in certain clinical work processes. Though ethical, regulatory, and legal issues arise, AI will contribute remarkably to revolutionizing current disease diagnostic patterns and generate a significant clinical impact in the near future.