Retinal microvasculature measured through fundus images holds predictive value in screening heart attacks and strokes. Moreover, the association between the neurosensory retina structure and neurodegenerative diseases has been well-established.
The thinning of retinal layers as measured in OCT scans can be used as a potential biomarker for the early detection of cognitive impairment in patients with Parkinson’s disease.
Blood-haemoglobin levels are predicted from retinal images and patient metadata using deep learning on data obtained from the UK Biobank.
Gender, smoking status, systolic blood pressure and major cardiac events are predicted through training a deep learning model on retinal fundus images.