Artificial intelligence system to detect eye and cardiovascular diseases

Summary Using the advanced deep learning system that has been developed and validated by the team as a basis, this project brings together medical research institutes, technical developers, industry, consumer organisations, government policy and service providers to develop, translate and prove an all-in-one AI system (A-Eye) that aims to create innovative solutions for multiple health disciplines and needs, including an opportunistic screening model, diagnosis standardisation and a cross-disciplinary model of risk prediction for cardiovascular diseases.

Project Details

Retinal photography gives us the ability to visualise the retina, optic disc, macula and blood vessels. It is the most important, low-cost and non-invasive diagnostic tool for common eye diseases such as glaucoma, diabetic retinopathy (DR), and age-related macular degeneration (AMD). The retinal microvasculature is a window to the body’s overall microvascular health. Subtle retinal microvascular changes have been identified as effective biomarkers to predict the risk of cardiovascular disease (CVD), which accounts for 14% of Australia’s total burden of disease. Currently, the interpretation of retinal imaging relies heavily on the subjective assessment of trained professionals.

Artificial intelligence (AI) offers the ability to revolutionise highly image driven medical specialities, including ophthalmology. Our research team has developed an AI system for the classification of the three aforementioned eye diseases and risk prediction for CVD. This technology has been validated and published in JAMA, Diabetes Care, Ophthalmology and JAMA Ophthalmology. Our AI system provides an opportunity to integrate AI technology with existing infrastructure to deliver an accurate, visually interpretable and user-friendly solution.

Using the advanced deep learning system that has been developed and validated by the team as a basis, this project brings together medical research institutes, technical developers, industry, consumer organisations, government policy and service providers to develop, translate and prove an all-in-one AI system (A-Eye) that aims to create innovative solutions for multiple health disciplines and needs, including an opportunistic screening model, diagnosis standardisation and a cross-disciplinary model of risk prediction for cardiovascular diseases. An “Eye and Systemic Disease AI Open Platform” will be created to enable resource sharing and workforce development to maximize the impact of AI on linking ocular imaging and systemic diseases.

This project is funded by MRFF Artificial intelligence research in health – Grant from 2020-2023.