
ASU and Mayo Clinic will study whether retinal images add value to Alzheimer’s blood markers. The project is research, not a routine screening test.

On October 28, 2026, Arizona State University announced a five-year, $3.2 million NIH-funded project led by professor Yalin Wang and Mayo Clinic neuro-ophthalmologist Oana M. Dumitrascu. The team will test whether improved retinal images combined with blood biomarkers can help identify Alzheimer’s risk before symptoms appear.
According to ASU News, the central question is whether retinal information adds something useful to blood-biomarker results. The researchers are especially interested in situations where blood results fall into an uncertain middle range.
Wang said blood biomarkers are already strong. The team’s aim is to find out whether patterns in retinal images can supply extra information when those blood results are less clear. This is a study of the value of combining information, rather than a claim that one test should replace another.
The ASU announcement describes the intended role as flagging risk before symptoms appear. It does not propose replacing brain scans or spinal-fluid tests or blood tests. In this context, “preclinical” refers to the period before noticeable symptoms.
That distinction helps define what a useful result would look like. The research needs to establish whether the retinal component contributes meaningful information beyond the blood markers. Simply collecting both types of data would not answer that question.
The planned approach has several parts:
Wang’s rationale is that the retina is biologically linked to the central nervous system and accessible for imaging. That makes it a candidate source of information about brain health. The project will investigate whether the proposed link translates into useful risk assessment.
The image-processing step is itself a research problem. ASU notes that AI-based image improvement could remove small blood-vessel details or introduce artificial ones. Either change could distort the signal the researchers are trying to study.
A picture that looks clearer is not necessarily a picture that preserves the right information. For this project, keeping the underlying retinal details intact matters more than producing a visually polished image. The model’s task is to analyze biological patterns rather than artifacts created during processing.
Collaborators include Mayo Clinic, Banner Alzheimer’s Institute, Banner Sun Health Research Institute and the Translational Genomics Research Institute (TGen). ASU says the team plans to include images collected through TGen’s mobile brain health lab in Arizona communities.
These community-collected images will let researchers examine pictures gathered outside ideal imaging conditions. That is a different challenge from working only with carefully controlled images. It brings image quality and the setting where pictures are collected into the research question.
Separately, Mayo Clinic describes Dumitrascu’s related retinal-biomarker work as entering a phase focused on validation and scale. Its account says she is planning prospective clinical testing in urban and rural settings, including mobile-care settings in Arizona.
The immediate relevance for adults over 60 is understanding where this research sits within memory care. A project looking for risk before symptoms asks a different question from a specialist evaluating cognitive problems that are already measurable.
A UK professional-guidance summary from NIHR Dementia Researcher describes plasma p-tau217 as part of a full specialist assessment when Alzheimer’s is suspected in someone with measurable cognitive problems. It is not described as a stand-alone test or as testing for people without symptoms outside research.
Psychiatric Times also describes the Alzheimer’s Association guideline as focused on blood biomarkers in the diagnostic workup of suspected Alzheimer’s within specialized care. Those recommendations concern clinical assessment rather than general population screening.
For someone noticing changes in memory or thinking, the practical step is to discuss those changes with a qualified healthcare professional. Test selection belongs within an individualized evaluation. FitBrainLab’s guide to how cognitive assessments and biomarkers fit together provides context for that discussion.
For people following the news without symptoms, the more useful question is what the research has actually demonstrated. FitBrainLab’s coverage of testing decisions before symptoms appear addresses that separate issue. Participation in research and routine clinical testing should not be treated as interchangeable.
According to Mayo Clinic, two peer-reviewed papers published in 2024 reported 85% accuracy for AI-analyzed retinal scans in identifying brain changes before noticeable cognitive deficits. That figure belongs to the earlier work described by Mayo. It is not the accuracy of the newly announced combined retinal-and-blood project or an established measure of real-world screening performance.
The ASU announcement reports a research plan, not completed results showing that the combined model identifies individual people at risk. Funding supports the effort to answer the question. It does not settle the question.
Wang cautioned that the project is research rather than a diagnostic test ready for routine optometry use. ASU says clinical use would require further validation and trials followed by regulatory review. The five-year research period is not a promised date for public availability.
The distinction between risk and diagnosis also matters. The project’s stated aim is to flag preclinical risk. Its announcement does not establish that a person’s retinal image can predict whether they will develop symptoms.
As future findings appear, several questions will help readers assess their meaning:
These are measures of progress that fit the project’s actual goals. They are more useful than treating the existence of an AI model as proof of a clinical advance.
Why does this development matter for cognitive health over the long term? It brings together questions that need to be answered together: what the retina contributes, how blood markers fit alongside it and whether the combined approach works with community-collected images.
A useful outcome would be evidence that clarifies where retinal information belongs in risk assessment. A finding that it adds little would also help define the technology’s limits. Either result would be more informative than assuming that an accessible image must be a useful screening tool.
For older adults, the long-term value would be assessment grounded in evidence from settings closer to everyday care. The strongest contribution this project can make is a clearer answer about whether adding retinal information genuinely improves the understanding of Alzheimer’s risk.
Questions about retinal screening fall to readers and their healthcare teams, while FitBrainLab takes on the editorial task of explaining the evidence. Its reporting separates early findings from established knowledge for adults 60+ who want clear information about memory and cognitive aging without alarm or medical promises.
Follow FitBrainLab for research-led insights on memory, focus, brain aging, nutrition and mental fitness after 60. Stay connected for new articles, practical guidance and ideas for a sharper, more engaged life.



Build habits that support memory, focus and a curious, connected life.
Read the Blog