Objective:

To develop a practical screening tool for Alzheimer's disease (AD) using label-free tri-spectral retinal imaging.

Key Findings:

• AD eyes showed increased blue reflectance compared to controls in the B/G ratiometric map.
• The machine learning model achieved an AUC of 0.83 in cross-validation and 0.91 in an independent test set.

Interpretation:

Tri-spectral imaging provides a non-invasive, scalable method for AD screening that integrates into existing ophthalmic workflows.

Limitations:

• The study is limited by its small sample size.
• Further validation with larger datasets is necessary before clinical application.

Conclusion:

Tri-spectral imaging holds promise for AD screening but requires more extensive research for clinical implementation.