Artificial intelligence-based retinal imaging for brain health assessment: a scoping review

Brain health is an emerging and rapidly expanding interdisciplinary field that requires optimisation across the life course of a person. The retina serves as an optically accessible window for studying brain health and related neurodegenerative and cerebrovascular diseases, which impose substantial global burdens. With advances in artificial intelligence (AI) and retinal imaging, the field of oculomics-defined as the identification of ocular imaging-based markers of systemic diseases-has grown considerably and shows promising applications in brain health. In this scoping review, we propose a potential ecosystem to deploy AI-based retinal imaging for assessing brain health and related neurodegenerative and cerebrovascular diseases. Overall, AI-based retinal imaging shows strong potential; however, further translation into clinical practice requires developing standardised benchmark datasets, enhancing model transparency, advancing interdisciplinary collaborations on clinical implementation, conducting rigorous validation in real-world populations, integrating models into clinical workflow, conducting cost-effectiveness analysis, establishing viable business models, and developing regulatory standards.