Label-free hyperspectral imaging and deep-learning prediction of retinal amyloid β-protein and phosphorylated tau

Hyperspectral retinal imaging in Alzheimer's disease and age-related macular degeneration: a review

While Alzheimer's disease and other neurodegenerative diseases have traditionally been viewed as brain disorders, there is growing evidence indicating their manifestation in the eyes as well. The retina, being a developmental extension of the brain, represents the only part of the central nervous system that can be noninvasively imaged at a high spatial resolution. The discovery of the specific pathological hallmarks of Alzheimer's disease in the retina of patients holds great promise for disease diagnosis and monitoring, particularly in the early stages where disease progression can potentially be slowed. Among various retinal imaging methods, hyperspectral imaging has garnered significant attention in this field. It offers a label-free approach to detect disease biomarkers, making it especially valuable for large-scale population screening efforts. In this review, we discuss recent advances in the field and outline the current bottlenecks and enabling technologies that could propel this field toward clinical translation.

Retinal ganglion cell vulnerability to pathogenic tau in Alzheimer's disease

Accumulation of pathological tau isoforms, especially hyperphosphorylated tau at serine 396 (pS396-tau) and tau oligomers, has been demonstrated in the retinas of patients with mild cognitive impairment (MCI) and Alzheimer's disease (AD). Previous studies have noted a decrease in retinal ganglion cells (RGCs) in AD patients, but the presence and impact of pathological tau isoforms in RGCs and RGC integrity, particularly in early AD stages, have not been explored. To investigate this, we examined retinal superior temporal cross-sections from 25 patients with MCI (due to AD) or AD dementia and 16 cognitively normal (CN) controls, matched for age and gender. We utilized the RGC marker ribonucleic acid binding protein with multiple splicing (RBPMS) and Nissl staining to assess neuronal density in the ganglion cell layer (GCL). Our study found that hypertrophic RGCs containing pS396-tau and T22-positive tau oligomers were more frequently observed in MCI and AD patients compared to CN subjects. Quantitative analyses indicated a decline in RGC integrity, with 46-55% and 55-56% reductions of RBPMS+ RGCs (P<0.01) and Nissl+ GCL neurons (P<0.01-0.001), respectively, in MCI and AD patients. This decrease in RGC count was accompanied by increases in necroptotic-like morphology and the cleaved caspase-3 apoptotic marker in RGCs of AD patients. Furthermore, there was a 2.1 to 3.1-fold increase (P<0.05-0.0001) in pS396-tau-laden RGCs in MCI and AD patients, with a greater abundance observed in individuals with higher Braak stages (V-VI), more severe clinical dementia ratings (CDR=3), and lower mini-mental state examination (MMSE) scores. Strong correlations were noted between the decline in RGCs and the total amount of retinal pS396-tau and pS396-tau+ RGCs, with pS396-tau+ RGC counts correlating significantly with brain neurofibrillary tangle scores (r= 0.71, P= 0.0001), Braak stage (r= 0.65, P= 0.0009), and MMSE scores (r= -0.76, P= 0.0004). These findings suggest that retinal tauopathy, characterized by pS396-tau and oligomeric tau in hypertrophic RGCs, is associated with and may contribute to RGC degeneration in AD. Future research should validate these findings in larger cohorts and explore noninvasive retinal imaging techniques that target tau pathology in RGCs to improve AD detection and monitor disease progression.

Rationale and design of the BeyeOMARKER study: prospective evaluation of blood- and eye-based biomarkers for early detection of Alzheimer's disease pathology in the eye clinic

BACKGROUND

Alzheimer's disease (AD) is a common, complex and multifactorial disease that may require screening across multiple routes of referral to enable early detection and subsequent future implementation of tailored interventions. Blood- and eye-based biomarkers show promise as low-cost, scalable and patient-friendly tools for early AD detection given their ability to provide information on AD pathophysiological changes and manifestations in the retina, respectively. Eye clinics provide an intriguing real-world proof-of-concept setting to evaluate the performance of these potential AD screening tools given the intricate connections between the eye and brain, presumed enrichment for AD pathology in the aging population with eye disorders, and the potential for an accelerated diagnostic pathway for under-recognized patient groups.

METHODS

The BeyeOMARKER study is a prospective, observational, longitudinal cohort study aiming to include individuals visiting an eye-clinic. Inclusion criteria entail being ≥ 50 years old and having no prior dementia diagnosis. Excluded eye-conditions include traumatic insults, superficial inflammation, and conditions in surrounding structures of the eye that are not engaged in vision. The BeyeOMARKER cohort (n = 700) will undergo blood collection to assess plasma p-tau217 levels and a brief cognitive screening at the eye clinic. All participants will subsequently be invited for annual longitudinal follow-up including remotely administered cognitive screening and questionnaires. The BeyeOMARKER + cohort (n = 150), consisting of 100 plasma p-tau217 positive participants and 50 matched negative controls selected from the BeyeOMARKER cohort, will additionally undergo Aβ-PET and tau-PET, MRI, retinal imaging including hyperspectral imaging (primary), widefield imaging, optical coherence tomography (OCT) and OCT-Angiography (secondary), and cognitive and cortical vision assessments.

RESULTS

We aim to implement the current protocol between April 2024 until March 2027. Primary outcomes include the performance of plasma p-tau217 and hyperspectral retinal imaging to detect AD pathology (using Aβ- and tau-PET visual read as reference standard) and to detect cognitive decline. Initial follow-up is ~ 2 years but may be extended with additional funding.

CONCLUSIONS

We envision that the BeyeOMARKER study will demonstrate the feasibility of early AD detection based on blood- and eye-based biomarkers in alternative screening settings, and will improve our understanding of the eye-brain connection.

TRIAL REGISTRATION

The BeyeOMARKER study (Eudamed CIV ID: CIV-NL-23-09-044086; registration date: 19th of March 2024) is approved by the ethical review board of the Amsterdam UMC.

Identification of retinal oligomeric, citrullinated, and other tau isoforms in early and advanced AD and relations to disease status

This study investigates various pathological tau isoforms in the retina of individuals with early and advanced Alzheimer's disease (AD), exploring their connection with disease status. Retinal cross-sections from predefined superior-temporal and inferior-temporal subregions and corresponding brains from neuropathologically confirmed AD patients with a clinical diagnosis of either mild cognitive impairment (MCI) or dementia (n = 45) were compared with retinas from age- and sex-matched individuals with normal cognition (n = 30) and non-AD dementia (n = 4). Retinal tau isoforms, including tau tangles, paired helical filament of tau (PHF-tau), oligomeric-tau (Oligo-tau), hyperphosphorylated-tau (p-tau), and citrullinated-tau (Cit-tau), were stereologically analyzed by immunohistochemistry and Nanostring GeoMx digital spatial profiling, and correlated with clinical and neuropathological outcomes. Our data indicated significant increases in various AD-related pretangle tau isoforms, especially p-tau (AT8, 2.9-fold, pS396-tau, 2.6-fold), Cit-tau at arginine residue 209 (CitR209-tau; 4.1-fold), and Oligo-tau (T22+, 9.2-fold), as well as pretangle and mature tau tangle forms like MC-1-positive (1.8-fold) and PHF-tau (2.3-fold), in AD compared to control retinas. MCI retinas also exhibited substantial increases in Oligo-tau (5.2-fold), CitR209-tau (3.5-fold), and pS396-tau (2.2-fold). Nanostring GeoMx analysis confirmed elevated retinal p-tau at epitopes: Ser214 (2.3-fold), Ser396 (2.6-fold), Ser404 (2.4-fold), and Thr231 (1.8-fold), particularly in MCI patients. Strong associations were found between retinal tau isoforms versus brain pathology and cognitive status: a) retinal Oligo-tau vs. Braak stage, neurofibrillary tangles (NFTs), and CDR cognitive scores (ρ = 0.63-0.71), b) retinal PHF-tau vs. neuropil threads (NTs) and ABC scores (ρ = 0.69-0.71), and c) retinal pS396-tau vs. NTs, NFTs, and ABC scores (ρ = 0.67-0.74). Notably, retinal Oligo-tau strongly correlated with retinal Aβ42 and arterial Aβ40 forms (r = 0.76-0.86). Overall, this study identifies and quantifies diverse retinal tau isoforms in MCI and AD patients, underscoring their link to brain pathology and cognition. These findings advocate for further exploration of retinal tauopathy biomarkers to facilitate AD detection and monitoring via noninvasive retinal imaging.

Alzheimer's disease pathophysiology in the Retina

The retina is an emerging CNS target for potential noninvasive diagnosis and tracking of Alzheimer's disease (AD). Studies have identified the pathological hallmarks of AD, including amyloid β-protein (Aβ) deposits and abnormal tau protein isoforms, in the retinas of AD patients and animal models. Moreover, structural and functional vascular abnormalities such as reduced blood flow, vascular Aβ deposition, and blood-retinal barrier damage, along with inflammation and neurodegeneration, have been described in retinas of patients with mild cognitive impairment and AD dementia. Histological, biochemical, and clinical studies have demonstrated that the nature and severity of AD pathologies in the retina and brain correspond. Proteomics analysis revealed a similar pattern of dysregulated proteins and biological pathways in the retina and brain of AD patients, with enhanced inflammatory and neurodegenerative processes, impaired oxidative-phosphorylation, and mitochondrial dysfunction. Notably, investigational imaging technologies can now detect AD-specific amyloid deposits, as well as vasculopathy and neurodegeneration in the retina of living AD patients, suggesting alterations at different disease stages and links to brain pathology. Current and exploratory ophthalmic imaging modalities, such as optical coherence tomography (OCT), OCT-angiography, confocal scanning laser ophthalmoscopy, and hyperspectral imaging, may offer promise in the clinical assessment of AD. However, further research is needed to deepen our understanding of AD's impact on the retina and its progression. To advance this field, future studies require replication in larger and diverse cohorts with confirmed AD biomarkers and standardized retinal imaging techniques. This will validate potential retinal biomarkers for AD, aiding in early screening and monitoring.

Hyperspectral retinal imaging biomarkers of ocular and systemic diseases

Hyperspectral imaging is a frontier in the field of medical imaging technology. It enables the simultaneous collection of spectroscopic and spatial data. Structural and physiological information encoded in these data can be used to identify and localise typically elusive biomarkers. Studies of retinal hyperspectral imaging have provided novel insights into disease pathophysiology and new ways of non-invasive diagnosis and monitoring of retinal and systemic diseases. This review provides a concise overview of recent advances in retinal hyperspectral imaging.

Predicting the cognitive impairment with multimodal ophthalmic imaging and artificial neural network for community screening

BACKGROUND/AIMS

To investigate the comprehensive prediction ability for cognitive impairment in a general elder population using the combination of the multimodal ophthalmic imaging and artificial neural networks.

METHODS

Patients with cognitive impairment and cognitively healthy individuals were recruited. All subjects underwent medical history, blood pressure measurement, the Montreal Cognitive Assessment, medical optometry, intraocular pressure and custom-built multimodal ophthalmic imaging, which integrated pupillary light reaction, multispectral imaging, laser speckle contrast imaging and retinal oximetry. Multidimensional parameters were analysed by Student's t-test. Logistic regression analysis and back-propagation neural network (BPNN) were used to identify the predictive capability for cognitive impairment.

RESULTS

This study included 104 cognitive impairment patients (61.5% female; mean (SD) age, 68.3 (9.4) years), and 94 cognitively healthy age-matched and sex-matched subjects (56.4% female; mean (SD) age, 65.9 (7.6) years). The variation of most parameters including decreased pupil constriction amplitude (CA), relative CA, average constriction velocity, venous diameter, venous blood flow and increased centred retinal reflectance in 548 nm (RC548) in cognitive impairment was consistent with previous studies while the reduced flow acceleration index and oxygen metabolism were reported for the first time. Compared with the logistic regression model, BPNN had better predictive performance (accuracy: 0.91 vs 0.69; sensitivity: 93.3% vs 61.70%; specificity: 90.0% vs 68.66%).

CONCLUSIONS

This study demonstrates retinal spectral signature alteration, neurodegeneration and angiopathy occur concurrently in cognitive impairment. The combination of multimodal ophthalmic imaging and BPNN can be a useful tool for predicting cognitive impairment with high performance for community screening.

Identification of retinal tau oligomers, citrullinated tau, and other tau isoforms in early and advanced AD and relations to disease status

Importance

This study identifies and quantifies diverse pathological tau isoforms in the retina of both early and advanced-stage Alzheimer's disease (AD) and determines their relationship with disease status.

Objective

A case-control study was conducted to investigate the accumulation of retinal neurofibrillary tangles (NFTs), paired helical filament (PHF)-tau, oligomeric tau (oligo-tau), hyperphosphorylated tau (p-tau), and citrullinated tau (Cit-tau) in relation to the respective brain pathology and cognitive dysfunction in mild cognitively impaired (MCI) and AD dementia patients versus normal cognition (NC) controls.

Design setting and participants

Eyes and brains from donors diagnosed with AD, MCI (due to AD), and NC were collected (n=75 in total), along with clinical and neuropathological data. Brain and retinal cross-sections-in predefined superior-temporal and inferior-temporal (ST/IT) subregions-were subjected to histopathology analysis or Nanostring GeoMx digital spatial profiling.

Main outcomes and measure

Retinal burden of NFTs (pretangles and mature tangles), PHF-tau, p-tau, oligo-tau, and Cit-tau was assessed in MCI and AD versus NC retinas. Pairwise correlations revealed associations between retinal and brain parameters and cognitive status.

Results

Increased retinal NFTs (1.8-fold, p=0.0494), PHF-tau (2.3-fold, p<0.0001), oligo-tau (9.1-fold, p<0.0001), CitR 209 -tau (4.3-fold, p<0.0001), pSer202/Thr205-tau (AT8; 4.1-fold, p<0.0001), and pSer396-tau (2.8-fold, p=0.0015) were detected in AD patients. Retinas from MCI patients showed significant increases in NFTs (2.0-fold, p=0.0444), CitR 209 -tau (3.5-fold, p=0.0201), pSer396-tau (2.6-fold, p=0.0409), and, moreover, oligo-tau (5.8-fold, p=0.0045). Nanostring GeoMx quantification demonstrated upregulated retinal p-tau levels in MCI patients at phosphorylation sites of Ser214 (2.3-fold, p=0.0060), Ser396 (1.8-fold, p=0.0052), Ser404 (2.4-fold, p=0.0018), and Thr231 (3.3-fold, p=0.0028). Strong correlations were found between retinal tau forms to paired-brain pathology and cognitive status: a) retinal oligo-tau vs. Braak stage (r=0.60, P=0.0002), b) retinal PHF-tau vs. ABC average score (r=0.64, P=0.0043), c) retinal pSer396-tau vs. brain NFTs (r=0.68, P<0.0001), and d) retinal pSer202/Thr205-tau vs. MMSE scores (r= -0.77, P=0.0089).

Conclusions and Relevance

This study reveals increases in immature and mature retinal tau isoforms in MCI and AD patients, highlighting their relationship with brain pathology and cognition. The data provide strong incentive to further explore retinal tauopathy markers that may be useful for early detection and monitoring of AD staging through noninvasive retinal imaging.

Clearance of interstitial fluid (ISF) and CSF (CLIC) group-part of Vascular Professional Interest Area (PIA), updates in 2022-2023. Cerebrovascular disease and the failure of elimination of Amyloid-β from the brain and retina with age and Alzheimer's disease: Opportunities for therapy

This editorial summarizes advances from the Clearance of Interstitial Fluid and Cerebrospinal Fluid (CLIC) group, within the Vascular Professional Interest Area (PIA) of the Alzheimer's Association International Society to Advance Alzheimer's Research and Treatment (ISTAART). The overarching objectives of the CLIC group are to: (1) understand the age-related physiology changes that underlie impaired clearance of interstitial fluid (ISF) and cerebrospinal fluid (CSF) (CLIC); (2) understand the cellular and molecular mechanisms underlying intramural periarterial drainage (IPAD) in the brain; (3) establish novel diagnostic tests for Alzheimer's disease (AD), cerebral amyloid angiopathy (CAA), retinal amyloid vasculopathy, amyloid-related imaging abnormalities (ARIA) of spontaneous and iatrogenic CAA-related inflammation (CAA-ri), and vasomotion; and (4) establish novel therapies that facilitate IPAD to eliminate amyloid β (Aβ) from the aging brain and retina, to prevent or reduce AD and CAA pathology and ARIA side events associated with AD immunotherapy.

Hyperspectral Retinal Imaging as a Non-Invasive Marker to Determine Brain Amyloid Status

Background

As an extension of the central nervous system (CNS), the retina shares many similarities with the brain and can manifest signs of various neurological diseases, including Alzheimer's disease (AD).

Objective

To investigate the retinal spectral features and develop a classification model to differentiate individuals with different brain amyloid levels.

Methods

Sixty-six participants with varying brain amyloid-β protein levels were non-invasively imaged using a hyperspectral retinal camera in the wavelength range of 450-900 nm in 5 nm steps. Multiple retina features from the central and superior views were selected and analyzed to identify their variability among individuals with different brain amyloid loads.

Results

The retinal reflectance spectra in the 450-585 nm wavelengths exhibited a significant difference in individuals with increasing brain amyloid. The retinal features in the superior view showed higher inter-subject variability. A classification model was trained to differentiate individuals with varying amyloid levels using the spectra of extracted retinal features. The performance of the spectral classification model was dependent upon retinal features and showed 0.758-0.879 accuracy, 0.718-0.909 sensitivity, 0.764-0.912 specificity, and 0.745-0.891 area under curve for the right eye.

Conclusions

This study highlights the spectral variation of retinal features associated with brain amyloid loads. It also demonstrates the feasibility of the retinal hyperspectral imaging technique as a potential method to identify individuals in the preclinical phase of AD as an inexpensive alternative to brain imaging.

Retina pathology as a target for biomarkers for Alzheimer's disease: Current status, ophthalmopathological background, challenges, and future directions

There is emerging evidence that amyloid beta protein (Aβ) and tau-related lesions in the retina are associated with Alzheimer's disease (AD). Aβ and hyperphosphorylated (p)-tau deposits have been described in the retina and were associated with small amyloid spots visualized by in vivo imaging techniques as well as degeneration of the retina. These changes correlate with brain amyloid deposition as determined by histological quantification, positron emission tomography (PET) or clinical diagnosis of AD. However, the literature is not coherent on these histopathological and in vivo imaging findings. One important reason for this is the variability in the methods and the interpretation of findings across different studies. In this perspective, we indicate the critical methodological deviations among different groups and suggest a roadmap moving forward on how to harmonize (i) histopathologic examination of retinal tissue; (ii) in vivo imaging among different methods, devices, and interpretation algorithms; and (iii) inclusion/exclusion criteria for studies aiming at retinal biomarker validation.

Retinal arterial Aβ40 deposition is linked with tight junction loss and cerebral amyloid angiopathy in MCI and AD patients

INTRODUCTION

Vascular amyloid beta (Aβ) protein deposits were detected in retinas of mild cognitively impaired (MCI) and Alzheimer's disease (AD) patients. We tested the hypothesis that the retinal vascular tight junctions (TJs) were compromised and linked to disease status.

METHODS

TJ components and Aβ expression in capillaries and larger blood vessels were determined in post mortem retinas from 34 MCI or AD patients and 27 cognitively normal controls and correlated with neuropathology.

RESULTS

Severe decreases in retinal vascular zonula occludens-1 (ZO-1) and claudin-5 correlating with abundant arteriolar Aβ40 deposition were identified in MCI and AD patients. Retinal claudin-5 deficiency was closely associated with cerebral amyloid angiopathy, whereas ZO-1 defects correlated with cerebral pathology and cognitive deficits.

DISCUSSION

We uncovered deficiencies in blood-retinal barrier markers for potential retinal imaging targets of AD screening and monitoring. Intense retinal arteriolar Aβ40 deposition suggests a common pathogenic mechanism of failed Aβ clearance via intramural periarterial drainage.

mHealth hyperspectral learning for instantaneous spatiospectral imaging of hemodynamics

Hyperspectral imaging acquires data in both the spatial and frequency domains to offer abundant physical or biological information. However, conventional hyperspectral imaging has intrinsic limitations of bulky instruments, slow data acquisition rate, and spatiospectral trade-off. Here we introduce hyperspectral learning for snapshot hyperspectral imaging in which sampled hyperspectral data in a small subarea are incorporated into a learning algorithm to recover the hypercube. Hyperspectral learning exploits the idea that a photograph is more than merely a picture and contains detailed spectral information. A small sampling of hyperspectral data enables spectrally informed learning to recover a hypercube from a red-green-blue (RGB) image without complete hyperspectral measurements. Hyperspectral learning is capable of recovering full spectroscopic resolution in the hypercube, comparable to high spectral resolutions of scientific spectrometers. Hyperspectral learning also enables ultrafast dynamic imaging, leveraging ultraslow video recording in an off-the-shelf smartphone, given that a video comprises a time series of multiple RGB images. To demonstrate its versatility, an experimental model of vascular development is used to extract hemodynamic parameters via statistical and deep learning approaches. Subsequently, the hemodynamics of peripheral microcirculation is assessed at an ultrafast temporal resolution up to a millisecond, using a conventional smartphone camera. This spectrally informed learning method is analogous to compressed sensing; however, it further allows for reliable hypercube recovery and key feature extractions with a transparent learning algorithm. This learning-powered snapshot hyperspectral imaging method yields high spectral and temporal resolutions and eliminates the spatiospectral trade-off, offering simple hardware requirements and potential applications of various machine learning techniques.

Retinal pathological features and proteome signatures of Alzheimer's disease

Alzheimer's disease (AD) pathologies were discovered in the accessible neurosensory retina. However, their exact nature and topographical distribution, particularly in the early stages of functional impairment, and how they relate to disease progression in the brain remain largely unknown. To better understand the pathological features of AD in the retina, we conducted an extensive histopathological and biochemical investigation of postmortem retina and brain tissues from 86 human donors. Quantitative examination of superior and inferior temporal retinas from mild cognitive impairment (MCI) and AD patients compared to those with normal cognition (NC) revealed significant increases in amyloid β-protein (Aβ42) forms and novel intraneuronal Aβ oligomers (AβOi), which were closely associated with exacerbated retinal macrogliosis, microgliosis, and tissue atrophy. These pathologies were unevenly distributed across retinal layers and geometrical areas, with the inner layers and peripheral subregions exhibiting most pronounced accumulations in the MCI and AD versus NC retinas. While microgliosis was increased in the retina of these patients, the proportion of microglial cells engaging in Aβ uptake was reduced. Female AD patients exhibited higher levels of retinal microgliosis than males. Notably, retinal Aβ42, S100 calcium-binding protein B+ macrogliosis, and atrophy correlated with severity of brain Aβ pathology, tauopathy, and atrophy, and most retinal pathologies reflected Braak staging. All retinal biomarkers correlated with the cognitive scores, with retinal Aβ42, far-peripheral AβOi and microgliosis displaying the strongest correlations. Proteomic analysis of AD retinas revealed activation of specific inflammatory and neurodegenerative processes and inhibition of oxidative phosphorylation/mitochondrial, and photoreceptor-related pathways. This study identifies and maps retinopathy in MCI and AD patients, demonstrating the quantitative relationship with brain pathology and cognition, and may lead to reliable retinal biomarkers for noninvasive retinal screening and monitoring of AD.

Alteration of the MEK1/2–ERK1/2 Signaling Pathway in the Retina Associated with Age and Development of AMD-Like Retinopathy

Age-related macular degeneration (AMD) is a complex neurodegenerative disease and a major cause of irreversible visual impairment in patients in developed countries. Although age is the greatest risk factor in AMD, molecular mechanisms involved in AMD remain unknown. Growing evidence shows that dysregulation of MAPK signaling contributes to aging and neurodegenerative diseases; however, the information on the role of MAPK upregulation in these processes is controversial. ERK1 and ERK2 participate in the maintenance of proteostasis through the regulation of protein aggregation induced by the endoplasmic reticulum stress and other stress-mediated cell responses. To assess the contribution of alterations in the ERK1/2 signaling to the AMD development, we compared age-associated changes in the activity of ERK1/2 signaling pathway in the retina of Wistar rats (control) and OXYS rats that develop AMD-like retinopathy spontaneously. The activity of the ERK1/2 signaling increased during physiological aging in the retina of Wistar rats. The manifestation and progression of the AMD-like pathology in the retina of OXYS rats was accompanied by hyperphosphorylation of ERK1/2 and MEK1/2, the key kinases of the ERK1/2 signaling pathway. The progression of the AMD-like pathology was also associated with the ERK1/2-dependent tau protein hyperphosphorylation and increase in the ERK1/2-dependent phosphorylation of alpha B crystallin at Ser45 in the retina.