In Vivo Assessment of Retinal Biomarkers by Hyperspectral Imaging: Early Detection of Alzheimer's Disease

A noninvasive and cost-effective means to detect preclinical Alzheimer's disease (AD) and monitor disease progression would be invaluable. The retina is a developmental extension of the brain and has been viewed as a window to evaluate AD-related pathology. Cross-sectional studies have shown structural changes in the retina of AD patients that include thinning of the retinal nerve-fiber layer and changes in retinal vasculature. However, such changes do not manifest in early stages of the disease nor are they specific biomarkers for AD. Described herein is the utilization of our retinal hyperspectral imaging (rHSI) technique as a biomarker for identification of AD-related early pathological changes in the retina. Specifically, this account concerns the translation of our rHSI technique from animal models to human AD subjects. The underlying principle is Rayleigh light scattering, which is expected from low-order A beta aggregates present in early pathology. Recruitment was restricted to AD subjects (N= 19) and age-matched controls, with no family history of AD (N= 16). To limit the influence of skin pigmentation, subjects were restricted to those with skin pigmentation values of 2-3 on the Fitzpatrick scale. The largest spectral deviation from control subjects, rHSI signature, was obtained at the MCI stage with MMSE scores >= 22, suggesting higher sensitivity of this technique in early disease stages. The rHSI signature observed is unaffected by eye pathologies such as glaucoma and cataract. Age of the subjects minimally influenced the spectral signatures. The rHSI technique shows promise for detection of preclinical AD; it is conducted in a truly noninvasive manner, without application of an exogenous label, and is thus potentially suitable for population screening.