A healthcare centre in Hong Kong has developed a system that uses artificial intelligence (AI) technologies to predict the risk of Alzheimer’s disease by analysing retinal images, marking a pioneering global breakthrough.
The system non-invasively analyses images of the back of the eye, detecting early changes in the blood vessels and nerves of the retina that may be linked to Alzheimer’s disease, allowing potential at-risk cases to be identified years before clinical symptoms appear.
The Humansa Medical Center, a longevity-focused health facility, developed the system in collaboration with i-Cognitio Sciences, an ophthalmic technology company from the Chinese University of Hong Kong (CUHK).
The AI developed by i-Cognitio was validated using a large dataset, including nearly 13,000 retinal images from 648 patients diagnosed with Alzheimer’s, and over 3,000 individuals with normal cognition.
A study published in 2022 in the journal Lancet Digital Health showed that the AI technology developed to detect Alzheimer’s risk achieves an accuracy rate of 80 to 92 per cent in multi-ethnic populations across various countries.
Dementia affects around one-third of people over the age of 85 in the Asia-Pacific region, according to recent data. In Hong Kong, 10 per cent of the population over the age of 70 is affected by dementia, with Alzheimer’s disease accounting for more than half of diagnosed cases.
Scientific studies highlight that up to 45 per cent of dementia cases could be prevented or delayed through early intervention, underscoring the importance of early detection in mitigating the impact of dementia on the population.
However, conventional cognitive tests and structural brain imaging offer limited accuracy, while advanced methods such as amyloid PET scans and cerebrospinal fluid analysis are invasive and less accessible to the general population.
In this context, the study developed a deep learning-based model with an approach that promises to offer “a simple, low-cost, and low-effort solution” for identifying potential patients in community settings, providing suitable accuracy and sensitivity.
In an effort to highlight the relevance of AI in brain diagnostics and healthcare, Vincent Mok, director and founder of i-Cognitio, emphasised the key role of the retina as a window to the brain.
According to Mok, “through non-invasive retinal imaging, it is possible to identify changes in the blood vessels and nerves of the retina associated with Alzheimer’s disease.”
The expert explained that these retinal changes may appear 10 to 15 years before the clinical symptoms of Alzheimer’s become evident, representing a crucial window for early intervention.
