Label-Free Hyperspectral Imaging and Deep-Learning Prediction of Retinal Amyloid β-Protein and Phosphorylated Tau

Abstract

Alzheimers disease (AD) is a major risk for the aging population. The pathological hallmarks of AD--an abnormal deposition of amyloid {beta}-protein (A{beta}) and phosphorylated tau (pTau)--have been demonstrated in the retinas of AD patients, including in prodromal patients with mild cognitive impairment (MCI). A{beta} pathology, especially the accumulation of the amyloidogenic 42-residue long alloform (A{beta}42), is considered an early and specific sign of AD, and together with tauopathy, confirms AD diagnosis. To visualize retinal A{beta} and pTau, state-of-the-art methods use fluorescence. However, administering contrast agents complicates the imaging procedure. To address this problem, we developed a label-free hyperspectral imaging method to detect the spectral signatures of A{beta}42 and pS396-Tau and predicted their abundance in retinal cross sections. For the first time, we reported the spectral signature of pTau and demonstrated an accurate prediction of A{beta} and pTau distribution powered by deep learning. We expect our finding will lay the groundwork for label-free detection of AD at its very earliest roots. Significance StatementThe pathological hallmarks of Alzheimers disease (AD), amyloid {beta}-protein (A{beta}) and hyperphosphorylated (p)Tau protein have been characterized by a hyperspectral camera in terms of spectral signatures. The unique optical properties of the hallmark proteins on the broad visible light range enable label-free and high-resolution detection and virtual staining of abnormal deposition in the retina tissue, which will lay the groundwork for AD early diagnosis and AD development quantification.