Reviewing Current Approaches to Assessing Biological Age from Retinal Imaging
The retina is the only part of the nervous system readily and easily imaged at low cost. It contains layers of delicate structure and microvessels, all of which accumulates visually distinct changes and damage with advancing age and cellular dysfunction. Given the spread, packaging, and standardization of machine learning technologies on the one hand, and the development of an increasing variety of aging clocks to assess biological age on the other, it was only a matter of time before someone (or several someones) applied machine learning to retinal imagery in an attempt to produce a retinal aging clock. As a general rule, any sufficiently complex set of biological data can be used to produce a reasonably effective measure of biological age. The data contained in images of the retina is no exception.
Today's open access review paper provides a concise tour of present efforts to build retinal aging clocks. This part of the field is far less developed than is the case for epigenetic clocks and other omics clocks. Nonetheless, it is interesting, in large part because a view of the retina is in large part a view of the health of capillaries. Capillary density is known to decrease with age in tissues throughout the body, and the retina is one of the few locations in the body where one can obtain a cost-effective assessment of capillary density. Loss of capillaries means a reduced flow of blood into tissues, and consequent issues of many sorts. It may well be one of the more important aspects of degenerative aging.
Estimating biological age from retinal imaging: a scoping review
This study aimed to appraise existing research using retinal photography to develop biological ageing markers. We sought to determine the accuracy of retinal age prediction models, evaluate their ability to reflect age-related parameters and explore their clinical associations. This scoping review identified models which estimate chronological age from retinal images with moderate to high accuracy and identified several age-related associations.
Four models are currently available to estimate biological age from retinal images, all based on deep learning algorithms: 'Retinal Age', 'EyeAge', 'convolutional network-based model', and 'RetiAGE'. 'Retinal Age', 'EyeAge', and 'convolutional network-based model' were trained to predict numerical chronological age from retinal images, while 'RetiAGE' was trained to predict the probability of an individual being older than 65 years.
All models were trained and validated using a single dataset, predominantly comprising Caucasian or Asian populations. To enhance robustness, both 'EyeAge' and 'RetiAGE' underwent additional internal testing on previously unseen images from the training and validation cohort. For model testing and outcome assessment, the UK Biobank was used by three models: 'Retinal Age', 'EyeAge' and 'RetiAGE'. While the four identified models demonstrated comparable accuracy and performance, it is important to highlight inconsistent reporting of performance metrics, with some pertaining to validation performance, and others test performance. Consequently, the generalisability of these models is uncertain, warranting further work to assess their applicability across diverse populations.
Nevertheless, using retinal age models to predict mortality and morbidity carries significant clinical implications. A key finding from selected papers emphasises that accelerated ageing, calculated as retinal age gap (RAG), age acceleration or other indices, consistently correlates with mortality risk across three models. In addition, 'Retinal Age' and 'EyeAge' show associations with cardiovascular disease, while 'Retinal Age' and 'convolutional network-based model' show connections with the risk of diabetic retinopathy in patients with diabetes. These findings highlight the potential of retinal age as an informative tool for quantifying risk of mortality and cardiovascular morbidity. However, no clinical trials have yet explored the utility or feasibility of the models, a crucial aspect for determining their clinical relevance. Furthermore, factors associated with higher RAG, including glycaemic status, central obesity, and metabolic syndrome, suggest that RAG may provide valuable insight into lifestyle habits and traits that accelerate ageing.