Retinal Imaging In Neurodegeneration is an important component in the neurobiology of neurodegenerative diseases. This page provides detailed information about its structure, function, and role in disease processes.
The retina is an embryological extension of the central nervous system, sharing developmental origin, cellular composition, and vascular properties with the brain. This anatomical relationship has established retinal imaging as a promising non-invasive approach for detecting and monitoring neurodegenerative diseases — a concept encapsulated by the phrase "the eye as a window to the brain" (London et al., 2013). Optical coherence tomography (OCT), OCT angiography (OCTA), and fundus imaging can quantify structural thinning of retinal neuronal layers, vascular density changes, and even protein deposits that mirror pathological processes in the brain (Wu et al., 2025).
In [Alzheimer's disease[/diseases/alzheimers, retinal nerve fiber layer (RNFL) and ganglion cell layer (GCL) thinning correlate with [hippocampal] volume loss and [Amyloid-Beta[/proteins/Amyloid-Beta burden. In [Parkinson's disease[/diseases/parkinsons, inner retinal layer thinning reflects [dopaminergic neurodegeneration[/mechanisms/dopaminergic-neurodegeneration in the [substantia nigra[/brain-regions/substantia-nigra. In [multiple sclerosis[/diseases/multiple-sclerosis, RNFL thinning serves as an established biomarker for axonal damage. These findings, combined with the emergence of artificial intelligence-based retinal analysis achieving >98% diagnostic accuracy, position retinal imaging as a potentially transformative screening and monitoring tool for neurodegenerative diseases (Alber et al., 2024).
OCT is the primary retinal imaging technology applied to neurodegeneration research. It uses low-coherence interferometry to generate cross-sectional images of retinal layers with micrometer-level resolution:
| Parameter |
Spectral Domain OCT |
Swept Source OCT |
| Axial resolution |
5-7 μm |
5-8 μm |
| Scan speed |
20,000-85,000 A-scans/sec |
100,000-400,000 A-scans/sec |
| Wavelength |
~840 nm |
~1050 nm |
| Penetration depth |
Moderate |
Enhanced (choroid, optic nerve) |
| Clinical availability |
Widely available |
Increasingly available |
OCT enables automated segmentation and quantification of individual retinal layers:
- Retinal nerve fiber layer (RNFL): Axons of retinal ganglion cells; the most studied layer in neurodegeneration
- Ganglion cell layer (GCL): Cell bodies of retinal ganglion cells
- Inner plexiform layer (IPL): Synaptic connections between ganglion cells and bipolar/amacrine cells
- GCL-IPL complex (GC-IPL): Combined ganglion cell and inner plexiform layer; highly sensitive to neurodegeneration
- Inner nuclear layer (INL): Contains bipolar, Müller, horizontal, and amacrine cells
- Outer nuclear layer (ONL): Photoreceptor cell bodies
- Total macular thickness: Sum of all retinal layers at the macula
OCTA extends OCT capabilities by detecting blood flow through motion contrast algorithms, providing non-invasive visualization of retinal vasculature without dye injection:
- Superficial capillary plexus (SCP): Vessels in the GCL and RNFL
- Deep capillary plexus (DCP): Vessels at the inner nuclear/outer plexiform layer boundary
- Foveal avascular zone (FAZ): Central avascular region; enlargement indicates capillary dropout
- Vessel density (VD): Percentage of vascularized area; reduced in neurodegenerative diseases
- Vessel length density (VLD): Total vessel length per unit area
¶ Fundus Photography and Fluorescein Angiography
Traditional fundus photography and fluorescein angiography provide complementary structural and dynamic vascular information, though with lower resolution than OCT/OCTA. Widefield imaging can capture peripheral retinal changes not assessed by macular OCT.
An emerging modality, hyperspectral imaging captures reflected light across multiple wavelengths to detect [Amyloid-Beta[/proteins/Amyloid-Beta deposits in the retina. Retinal amyloid imaging in [Alzheimer's disease[/diseases/alzheimers patients has demonstrated curcumin-labeled amyloid plaques in the retina that correlate with brain amyloid burden (Koronyo et al., 2017).
[Alzheimer's disease[/diseases/alzheimers is the most extensively studied neurodegenerative condition in retinal imaging research:
Structural changes (OCT):
- RNFL thinning: Meta-analyses demonstrate significant peripapillary RNFL thinning in AD patients, particularly in the superior and inferior quadrants. The mean difference is approximately 5-10 μm compared to age-matched controls (den Haan et al., 2017)
- GC-IPL thinning: The ganglion cell-inner plexiform layer complex shows significant thinning, with some studies suggesting it is more sensitive than RNFL for detecting early AD pathology
- Macular volume reduction: Total macular volume is reduced in AD and [MCI[/diseases/mci
- MCI detection: Significant thinning of inferior/superior peripapillary RNFL and inner macular thickness in MCI suggests these as potential biomarkers for early neurodegeneration
Vascular changes (OCTA):
- Vessel density reduction: Decreased VD in both superficial and deep capillary plexuses in AD and MCI (Bulut et al., 2018)
- FAZ enlargement: The foveal avascular zone area is increased in AD, reflecting capillary loss
- Perfusion deficits: Vessel perfusion density is reduced, correlating with cerebral hypoperfusion patterns seen on [PET imaging[/diagnostics/pet-imaging
Multimodal integration: A 2025 study combining spectral domain OCT retinal parameters with [plasma biomarkers[/diagnostics/plasma-biomarkers achieved an area under the curve (AUC) of 0.97 for distinguishing amyloid-positive individuals from controls, demonstrating the power of multimodal approaches (Ashraf et al., 2025).
Retinal amyloid deposits: Post-mortem and in vivo studies have identified [Amyloid-Beta[/proteins/Amyloid-Beta plaques in the retinas of AD patients, suggesting the retina may harbor the same proteinopathy as the brain (Koronyo et al., 2017).
Retinal changes in [Parkinson's disease[/diseases/parkinsons reflect both retinal dopaminergic dysfunction and more general neurodegeneration:
Structural findings:
- Inner retinal thinning: Loss of [amacrine cells[/cell-types/dopaminergic-neurons-snpc (which are dopaminergic in the retina) and ganglion cells produces measurable IPL and GCL thinning
- RNFL thinning: Peripapillary RNFL is thinner in PD, particularly in the temporal quadrant, and correlates with disease duration and severity
- Foveal pit morphology: Changes in foveal architecture have been reported in early PD
Vascular findings (OCTA):
- Decreased vessel length density in central, inner, and full regions
- Reduced vessel perfusion density across all OCTA measurement zones
- Decreased FAZ circularity index
Correlation with brain pathology: Retinal structural alterations, particularly loss of amacrine and ganglion cells, correlate with dopaminergic neuron loss in the [substantia nigra[/brain-regions/substantia-nigra. Retinal thinning measured by OCT can serve as a biomarker for disease progression and severity (Murueta-Goyena et al., 2021).
[multiple sclerosis[/diseases/multiple-sclerosis represents the most clinically established application of retinal OCT in neurological disease:
- RNFL thinning: Well-documented in MS, particularly after optic neuritis episodes; used as a surrogate endpoint in clinical trials
- GCL thinning: Progressive GCL-IPL loss even without clinical optic neuritis indicates subclinical retinal neurodegeneration
- INL thickening: Microcystic macular edema in the inner nuclear layer may indicate [blood-brain barrier[/entities/blood-brain-barrier disruption and active inflammation
- Clinical integration: OCT is increasingly used in clinical MS practice for monitoring disease activity and treatment response
Retinal thinning has been documented in [Huntington's disease[/mechanisms/huntington-pathway:
- RNFL and macular thinning correlate with disease stage and CAG repeat length
- Temporal RNFL thinning is the most consistent finding
- Retinal changes may precede motor symptom onset in premanifest HD gene carriers
Retinal changes in [ALS[/diseases/als are increasingly recognized:
- Macular thinning and RNFL changes suggest that neurodegeneration extends beyond the motor system
- Retinal vascular changes may reflect broader neurovascular dysfunction
Retinal biomarkers are being explored in [FTD[/diseases/ftd:
- RNFL thinning patterns may differ from AD, potentially aiding differential diagnosis
- Retinal vascular changes are under investigation as early biomarkers
Deep learning has dramatically improved the diagnostic potential of retinal imaging for neurodegeneration:
- AD classification: CNNs applied to OCT images achieve average accuracy of 98.18% for raw images and 98.91% for segmented images in distinguishing AD from healthy controls (Wisely et al., 2024)
- Multi-disease screening: AI algorithms can simultaneously screen for multiple neurodegenerative conditions from a single retinal scan
- Risk prediction: Retinal AI models can predict future cognitive decline and dementia risk from baseline retinal images
- Fundus photography: Even standard fundus photographs, analyzed by deep learning, can predict cardiovascular and neurological risk factors relevant to neurodegeneration
¶ Advantages and Limitations
- Non-invasive: No radiation, no injection (for OCT/OCTA), minimal patient burden
- Cost-effective: OCT is widely available in ophthalmology clinics; far less expensive than [PET imaging[/diagnostics/pet-imaging or [MRI[/diagnostics/neuroimaging
- Rapid acquisition: OCT scans take seconds to acquire
- High resolution: Micrometer-level imaging of individual retinal layers
- Repeatability: High test-retest reliability enables longitudinal monitoring
- Widespread availability: OCT is standard equipment in ophthalmology; could enable population-level screening
- Accessible in primary care: Portable OCT devices are entering the market
- Small effect sizes: Group differences in retinal thickness between disease and control groups are small (typically 5-10 μm), limiting individual-level diagnostic precision
- Confounding conditions: Glaucoma, diabetic retinopathy, age-related macular degeneration, and other ophthalmic conditions produce similar retinal thinning
- Age-related changes: Normal aging produces RNFL and macular thinning that overlaps with neurodegenerative changes
- Lack of pathological specificity: RNFL thinning is common across many neurological conditions and cannot distinguish between them independently
- Standardization challenges: Different OCT devices, segmentation algorithms, and scan protocols limit cross-study comparisons
- Limited evidence for individual diagnosis: While group differences are robust, sensitivity and specificity for individual patient diagnosis remain insufficient for standalone use
¶ Clinical Recommendations and Future Directions
Retinal imaging is not yet recommended as a standalone diagnostic tool for neurodegenerative diseases but is recognized as a promising complementary biomarker. Current consensus recommendations suggest:
- Multimodal integration: Combining retinal OCT/OCTA with [plasma biomarkers[/diagnostics/plasma-biomarkers, [cognitive assessments[/diagnostics/cognitive-assessments, and clinical data maximizes diagnostic accuracy
- Longitudinal monitoring: Serial OCT measurements may be more informative than single time-point assessments for tracking disease progression
- Standardized protocols: Adoption of standardized OCT acquisition and analysis protocols is essential for clinical implementation
- Population screening: Retinal imaging's low cost and widespread availability make it a candidate for population-level screening in primary care settings, particularly when combined with AI analysis
- Clinical trial endpoints: Retinal OCT is being evaluated as a surrogate endpoint for neuroprotection trials
The study of Retinal Imaging In Neurodegeneration has evolved significantly over the past decades. Research in this area has revealed important insights into the underlying mechanisms of neurodegeneration and continues to drive therapeutic development.
Historical context and key discoveries in this field have shaped our current understanding and will continue to guide future research directions.
- DOI
- DOI
- DOI
- DOI
- DOI
- PubMed
- DOI
- DOI
- DOI
- PubMed
- PubMed