Retinal spectral intensity ratios

1. Retinal spectral intensity ratios predicts Alzheimer's disease in silico (accuracy 82%, sensitivity 87%, AUC 0.89).

   “The XGBoost model incorporating spectral intensity ratios achieved high classification performance, with an accuracy of 82%, sensitivity of 87%, and an area under the ROC curve (AUC) of 0.89.”

   In silicoclassifying AD vs healthy subjects

   Reflectance Spectral Images of the Retina as a Biosensor of Alzheimer's Disease

   AAIC 2025 abstract·8 Jan 2026

2. Alzheimer's disease is associated with retinal spectral intensity ratios in humans (p < 0.001).

   “Significant differences in spectral intensity ratios between AD and healthy subjects were observed ( p < 0.001).”

   HumansAD vs age-matched healthy subjects

   Reflectance Spectral Images of the Retina as a Biosensor of Alzheimer's Disease

   AAIC 2025 abstract·8 Jan 2026

More on Alzheimer's disease