Summary: New research that used artificial intelligence to analyze normal eye images has opened up a cheap, non-invasive way to predict key risk factors for Alzheimer's disease decades before clinical symptoms appear. By training machine learning models on retinal images from more than 40,000 patients in a UK-based data bank, the research team mapped specific areas of the eye, such as the retinal vessels and optic nerve, to biological and lifestyle risk factors associated with Alzheimer's vulnerability.
The AI accurately predicted factors including natural sex, blood pressure, smoking, alcohol consumption, and insomnia. Because the retina acts as a direct extension of the central nervous system, these common, low-cost images act as an “integrated biological sensor” for a patient's cumulative neurological damage, opening an important window for early lifestyle and medical intervention before irreversible brain damage occurs.
Important Facts
- Ocular window: Retinal morphology (especially vessels and optic nerve) provides measurable indicators of neurovascular integrity, serving as an organic sensor of Alzheimer's vulnerability.
- The Great Trial of the Databank: A machine learning model was trained and validated using retinal images from more than 40,000 patients archived in a large database in the United Kingdom.
- Purpose of the Risk Map: AI successfully bypasses unreliable patient reports to correctly identify lifestyle and biological risks, including high blood pressure, smoking, alcohol use, and insomnia.
- A Decade-Early Intervention: Because Alzheimer's disease develops over decades, this low-cost test identifies patients at risk long before late-stage, irreversible brain damage occurs.
- All Locations & Costs: Unlike MRIs or less expensive PET scans, retinal imaging is already widely performed during routine eye exams, diabetes screenings, and glaucoma screenings.
Source: University of Florida
Often referred to as the “window to the soul,” the eyes may also provide clues about something less poetic but more important: brain health.
A new study of tens of thousands of patients has revealed that cheap, simple and routine imaging of the retina behind the eye can accurately predict many of the most common risk factors associated with developing Alzheimer's disease.
“We know that Alzheimer's disease develops over decades, but most diagnostic tools focus on late-stage disease when it's too late to intervene,” said Ruogu Fang, Ph.D., a professor of bioengineering at the University of Florida who led the new study. “By looking at novel biomarkers, such as retinal health, we offer new opportunities to identify patients at risk, offer appropriate tests and encourage them to develop a healthy lifestyle to reduce their risk.”
Fang and his collaborators, including UF's Adam Woods, Ph.D., and Meta researcher Yunchao Yang, Ph.D., published their findings on June 16 Journal of Alzheimer's Disease.
Many patients often take pictures of their eyes. Those with diabetes, glaucoma or cataracts will have more retinal photos taken over the years. Even a routine eye exam with prescription glasses can capture images. That all-in-one makes analyzing retinal images easier and less expensive compared to other, more expensive technologies like MRIs.
By using machine learning to analyze these retinal images from more than 40,000 patients in a patient data bank based in the United Kingdom, Fang's group was able to identify retinal regions associated with Alzheimer's risk factors, such as blood vessels and optic nerves.
“With the help of AI, we are now able to identify subtle retinal variations that were previously overlooked in thousands of studies, which may serve as reliable indicators of future disease risk,” said Seowung Leem, a doctoral student at UF and first author of the paper.
The AI model accurately predicted biological factors such as sex or blood pressure and lifestyle factors associated with Alzheimer's disease, such as smoking, alcohol use and insomnia. Although many of these factors are captured in patients' medical charts, those records are often incomplete. Others, like alcohol and smoking, rely on unreliable self-reporting.
Retinal imaging may therefore provide an alternative, more objective method for detecting these risk factors. Also, retinal images can capture damage that has accumulated over the years, which will vary between patients who share the same risk factors.
“Retinal morphology can provide quantifiable indicators of neurovascular integrity, which is critical to Alzheimer's disease risk,” said Fang, who is also a member of the McKnight Brain Institute. “In this sense, retinal imaging functions less as a diagnostic questionnaire and more as an integrated biological sensor of increased risk.”
Fang's group has already found that retinal imaging can detect active cases of Alzheimer's disease. But scientists now believe that the disease continues for many years, even decades. Early identification of risk factors can therefore better identify patients who can respond to earlier interventions – including preventive lifestyle changes, certain medications or brain training – before irreversible brain damage occurs.
Sponsorship: This work was supported in part by the National Science Foundation.
Important Questions Answered:
A: The retina is developmentally and anatomically an extension of the central nervous system; it is literally part of the brain tissue that is pushed into the eye during fetal development. Because they share the same embryological origin, the blood vessels and nerve fibers behind your eye mirror the microvasculature and neural pathways inside your cranium. Damage from systemic inflammation, poor sleep, or vascular disease is reflected in the delicate structure of the retina, making it a visual proxy for the hidden state of the brain.
A: Traditional charts that rely heavily on self-reports of lifestyle risk factors such as alcohol consumption, smoking habits, or insomnia levels, are inaccurately reported or underestimated by patients. In addition, the chart does not show the actual damage a certain way of life has caused. The AI model learns the physical, cumulative toll these habits have taken on the ocular tissues over the decades. It transforms a simple image into a targeted biological sensor that shows real-time neurovascular risk.
A: Early detection shifts the timeline from effective treatment to effective prevention. Because brain damage takes decades to manifest clinically, early detection allows patients to make protective lifestyle changes while the brain is still working hard. These include intensive cardiovascular control, special brain training regimens, intensive sleep treatments to promote detoxification, and potentially early-onset treatments that are more effective before neurological death spreads.
Editor's Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper is fully revised.
- More content has been added by our staff.
About this AI and Alzheimer's research news
Author: Eric Hamilton
Source: University of Florida
Contact person: Eric Hamilton – University of Florida
Image: Image posted in Neuroscience News
Actual research: Closed access.
“Prediction of Alzheimer's disease risk factors in retinal images by deep learning: Development and validation of biologically morphological associations in the UK Biobank” by Ruogu Fang et al. Journal of Alzheimer's Disease
DOI:10.1177/13872877261457650
Abstract
Prediction of Alzheimer's disease risk factors in retinal images by deep learning: Development and validation of biologically morphological associations in the UK Biobank.
It's the background
Systemic, metabolic, and lifestyle factors have established associations with Alzheimer's disease (AD) through epidemiologic and AD-specific biomarker studies. Whether color fundus imaging (CFP) contains retinal structural signatures consistent with these AD-related risk domains remains unclear.
The purpose
To determine if deep learning (DL) models can predict 12 AD-related risk factors from CFP and show the retinal structures underlying these predictions, thus testing whether CFP shows pathways to AD vulnerability.
Methods
Using 62,876 CFPs from 44,501 unique participants from the UK Biobank, DL models were trained to predict 12 factors related to AD pathology or events: 6 categories (gender, smoking, insomnia, economic status, alcohol use, depression) and 6 continuous (age, age at completion of body weight, diasy, diasy education). HbA1c). Model performance, model robustness, and saliency-based scores (CAM-Score) were evaluated and compared with retinal morphometry. Scores were also compared between AD cases (average 8.55 years before onset) and matched controls.
Results
The predictive performance of DL ranged from AUROC between 0.5654 and 0.9480 for categorical factors and R2between −0.0291 and 0.7620 due to continuous factors, which perform better than morphometry-based machine learning models. The saliency-based score consistently highlighted regions with biological significance, particularly the optic nerve head and retinal vasculature. It also corresponds to the current morphometric variation. Several saliency-based scores differed significantly between incident AD and matched controls, suggesting potential overlap between retinal correlates of AD-related risk factors and preclinical AD-related changes.
Conclusions
CFP encodes retinal signatures linked to AD risk factors. Although not diagnostic, retinal representations found in DL may reveal structural changes related to biological risk that indicate potential AD vulnerability.
