Retinal Artificial Intelligence‐based Model Identifies Non‐demented Elderly Subjects at Risk of Alzheimer's Disease
Bibliographic record
Abstract
BACKGROUND: RetinAD is a validated deep learning model for differentiating between Alzheimer's disease (AD) dementia and cognitively unimpaired subjects based on analyzing retinal photographs. Since certain AD-related retinal changes (e.g., microvasculopathy) may start to develop years to decades before the onset of cognitive symptoms, we hypothesized that RetinAD may also identify retinal microvasculopathy among non-demented elderly subjects. We aimed to compare measures of retinal vessel network between "positive" and "negative" cases as classified by RetinAD among elderly non-demented elderly subjects. METHOD: We recruited community subjects who were participants in the BEAT AD (Brain Health Education And Tailor-made Measures for Prevention of Alzheimer's Disease) service programme in Hong Kong. This programme invites non-demented community dwelling subjects (59-80 years old) with subjective cognitive decline (SCD). It assesses their cognitive performances using Montreal Cognitive Assessment-5 minutes (MoCA-5) and on their control in the modifiable risk factors of AD. It also provides tailor-made recommendation for the subjects of how to optimize those risk factors that are not well controlled. We obtained fundus pictures using the Topcon NW500 non-mydriatic retinal camera. We classified subjects into "positive" or "negative" using RetinAD. We conducted quantitative measurements of retinal vessels using the Singapore I Vessel Assessment (SIVA) software. RESULT: Among the 187 recruited subjects with SCD, 29 (15.5%) and 158 (84.5%) subjects were classified as "positive" and "negative", respectively. Subjects who were classified as "positive" were older (mean age 71.21 versus [vs] 67.59; p = <0.01) than those who were classified as "negative". There was no significant difference in MoCA scores between "positive" (22.79) and "negative" subjects (23.74, p = 0.28). Analysis of the retinal vessel network showed that "positive" subjects had a significantly higher branching coefficient arterioles (1.64 vs 1.49) and branching coefficient venules (1.49 vs 1.32) than that of "negative" subjects. The difference remained significant (p = 0.033) for the branching coefficient venules after being adjusted to age and mean arterial pressure. CONCLUSION: RetinAD identified non-demented elderly who had worse retinal microvasculopathy and biologically older brains. Findings suggested that RetinAD may be able to identify elderly subjects who are at risk of developing AD dementia in the future.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".