Retinal Imaging: A Window into Systemic and Neurological Health

Many thanks to our sponsor Esdebe who helped us prepare this research report.

Abstract

Retinal imaging, a non-invasive and readily accessible diagnostic modality, has emerged as a powerful tool for investigating a wide spectrum of systemic and neurological disorders. Beyond its traditional role in ophthalmology, advancements in retinal imaging technologies and analysis techniques are revealing its potential to provide early detection and monitoring capabilities for conditions such as Alzheimer’s disease, cardiovascular disease, multiple sclerosis, and diabetic retinopathy. This research report comprehensively reviews the current state-of-the-art in retinal imaging, encompassing various modalities like optical coherence tomography (OCT), fundus photography, adaptive optics imaging, and angiography. We explore the underlying principles of these techniques, their specific applications in detecting a range of diseases, and the biomarkers that are accessible through retinal assessment. Furthermore, we address the challenges and limitations inherent in retinal imaging, including image quality, patient variability, and the need for standardized protocols. Finally, we discuss future directions in the field, emphasizing the transformative potential of advanced imaging modalities, artificial intelligence (AI)-driven analysis, and multimodal imaging approaches to enhance diagnostic accuracy and predictive power, ultimately improving patient outcomes.

Many thanks to our sponsor Esdebe who helped us prepare this research report.

1. Introduction

The retina, a light-sensitive neural tissue lining the posterior aspect of the eye, is uniquely positioned as a directly observable extension of the central nervous system (CNS). Its accessibility via non-invasive imaging techniques has revolutionized ophthalmology and, increasingly, is impacting other medical fields. The microvasculature of the retina mirrors that of the brain, making it a valuable surrogate for studying cerebrovascular health. Furthermore, the retinal nerve fiber layer (RNFL) comprises unmyelinated axons of retinal ganglion cells, representing a window into neuronal integrity. Consequently, retinal imaging offers a unique opportunity to visualize and quantify microvascular and neural changes that may precede or accompany systemic and neurological diseases.

Traditional retinal imaging techniques, such as fundus photography and fluorescein angiography, have long been used for diagnosing and monitoring ocular diseases like diabetic retinopathy, age-related macular degeneration (AMD), and glaucoma [1, 2]. However, recent technological advancements have significantly expanded the scope and sensitivity of retinal imaging, enabling the detection of subtle changes indicative of systemic and neurological conditions. Optical coherence tomography (OCT), for example, provides high-resolution cross-sectional images of the retina, allowing for precise measurement of RNFL thickness, macular volume, and other structural parameters [3]. Furthermore, techniques like OCT angiography (OCTA) offer non-invasive visualization of retinal vasculature, revealing microvascular abnormalities that are relevant to both ocular and systemic diseases [4]. Adaptive optics (AO) imaging, including adaptive optics scanning laser ophthalmoscopy (AOSLO) and adaptive optics OCT (AO-OCT), further enhance image resolution, enabling visualization of individual cells and capillaries [5].

This research report aims to provide a comprehensive overview of retinal imaging as a powerful tool for investigating systemic and neurological health. We will delve into the various retinal imaging techniques, their underlying principles, and their specific applications in detecting and monitoring a range of diseases. We will also address the challenges and limitations associated with retinal imaging and explore future directions in the field, focusing on the integration of advanced imaging modalities and AI-driven analysis.

Many thanks to our sponsor Esdebe who helped us prepare this research report.

2. Retinal Imaging Modalities

Several distinct retinal imaging modalities are used in clinical and research settings. Each modality offers unique advantages and limitations in terms of resolution, penetration depth, and specific biomarkers that can be assessed. This section will discuss the principles and applications of the most commonly used techniques.

2.1. Fundus Photography

Fundus photography is a non-invasive technique that captures a color image of the retina. It is widely used for documenting retinal findings, such as optic disc cupping in glaucoma, drusen in AMD, and hemorrhages or exudates in diabetic retinopathy [6]. Fundus photography is relatively inexpensive and easy to perform, making it a valuable screening tool. Widefield fundus photography expands the field of view, allowing for visualization of the peripheral retina, which is particularly useful in detecting peripheral retinal lesions associated with diabetic retinopathy or retinal detachments [7].

2.2. Fluorescein Angiography (FA) and Indocyanine Green Angiography (ICGA)

FA and ICGA are invasive imaging techniques that involve injecting a dye (fluorescein or indocyanine green) intravenously and then capturing a series of images as the dye circulates through the retinal and choroidal vasculature, respectively. FA is primarily used to visualize retinal vascular abnormalities, such as neovascularization, leakage, and capillary non-perfusion [8]. ICGA, due to its ability to penetrate deeper into the choroid, is used to visualize choroidal vascular abnormalities, such as choroidal neovascularization and polypoidal choroidal vasculopathy [9]. While FA and ICGA provide valuable information about vascular function, they are associated with potential side effects, including allergic reactions and, rarely, more serious complications.

2.3. Optical Coherence Tomography (OCT)

OCT is a non-invasive imaging technique that uses light waves to create high-resolution, cross-sectional images of the retina. It operates on the principle of low-coherence interferometry, measuring the echo time delay and magnitude of backscattered light [3]. OCT allows for precise measurement of retinal layer thickness, including the RNFL, ganglion cell layer (GCL), and inner plexiform layer (IPL). Spectral-domain OCT (SD-OCT) and swept-source OCT (SS-OCT) are advanced versions of OCT that offer faster scanning speeds and higher resolution than time-domain OCT [10].

OCT is widely used in ophthalmology for diagnosing and monitoring a variety of retinal diseases, including glaucoma, AMD, and diabetic macular edema. In the context of neurological diseases, OCT is particularly useful for assessing RNFL thickness, which can be reduced in conditions like multiple sclerosis and Alzheimer’s disease [11, 12].

2.4. Optical Coherence Tomography Angiography (OCTA)

OCTA is a non-invasive imaging technique that uses OCT technology to visualize retinal and choroidal vasculature without the need for intravenous dye injection. OCTA relies on motion contrast algorithms to detect changes in the OCT signal caused by flowing blood cells [4]. OCTA provides detailed information about the density and morphology of retinal and choroidal capillaries, allowing for the detection of microvascular abnormalities, such as capillary non-perfusion, microaneurysms, and neovascularization. The ability to visualize and quantify microvasculature changes makes OCTA a valuable tool for assessing diseases like diabetic retinopathy, glaucoma, and AMD, and emerging evidence suggests utility in neurological conditions.

2.5. Adaptive Optics (AO) Imaging

AO imaging is a technology that corrects for optical aberrations in the eye, resulting in significantly improved image resolution. AO systems typically consist of a wavefront sensor that measures the aberrations and a deformable mirror that compensates for them in real-time. AO can be integrated with various imaging modalities, such as scanning laser ophthalmoscopy (AOSLO) and OCT (AO-OCT), to achieve cellular-level resolution [5]. AOSLO allows for visualization of individual photoreceptors, retinal pigment epithelial cells, and capillaries. AO-OCT combines the high resolution of AO with the cross-sectional imaging capabilities of OCT, enabling detailed visualization of retinal microstructure. AO imaging is a powerful research tool for studying retinal disease at the cellular level, and is beginning to enter clinical practice.

Many thanks to our sponsor Esdebe who helped us prepare this research report.

3. Retinal Imaging in Neurological Diseases

The retina, being a part of the CNS, shares structural and functional similarities with the brain. This connection makes retinal imaging a promising tool for investigating and monitoring neurological diseases. Several studies have demonstrated the utility of retinal imaging in detecting early changes associated with conditions such as Alzheimer’s disease, Parkinson’s disease, and multiple sclerosis.

3.1. Alzheimer’s Disease

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline and memory loss. Pathologically, AD is associated with the accumulation of amyloid plaques and neurofibrillary tangles in the brain [13]. Emerging evidence suggests that similar pathological changes may also occur in the retina. Specifically, amyloid-beta (Aβ) plaques have been detected in the retinas of AD patients using fluorescent amyloid binding dyes and hyperspectral imaging [14].

Several studies have shown that AD patients exhibit reduced RNFL thickness, particularly in the temporal quadrant, as measured by OCT [12, 15]. Additionally, decreased macular volume and ganglion cell-inner plexiform layer (GCIPL) thickness have also been reported in AD patients [16]. These structural changes in the retina may reflect neuronal loss and neurodegeneration occurring in the brain. OCTA studies have shown reduced retinal vascular density in AD patients, particularly in the superficial capillary plexus [17]. This may reflect the microvascular dysfunction and reduced cerebral blood flow that are known to occur in AD.

While these findings are promising, it is important to note that the sensitivity and specificity of retinal imaging for detecting AD are still being investigated. Further research is needed to determine the optimal retinal biomarkers and imaging protocols for early AD detection and risk stratification. The lack of standardized protocols and variations in patient populations also contribute to the challenges in interpreting and comparing results across different studies. Additionally, the influence of coexisting ocular diseases, such as glaucoma and AMD, on retinal structure and function must be carefully considered when using retinal imaging to assess AD risk.

3.2. Parkinson’s Disease

Parkinson’s disease (PD) is a neurodegenerative disorder characterized by motor symptoms such as tremor, rigidity, and bradykinesia. The underlying pathology involves the loss of dopaminergic neurons in the substantia nigra [18]. In addition to motor symptoms, PD patients often experience non-motor symptoms, including visual disturbances. Several studies have investigated the use of retinal imaging to detect changes associated with PD.

Similar to AD, PD patients have been shown to exhibit reduced RNFL thickness, particularly in the inferior quadrant, as measured by OCT [19]. Additionally, decreased macular volume and GCIPL thickness have also been reported in PD patients [20]. These structural changes in the retina may reflect the neurodegeneration that occurs in the brain. Dopamine plays a crucial role in retinal function, particularly in contrast sensitivity and color vision. The loss of dopaminergic neurons in PD may contribute to retinal dysfunction. Some studies using pattern electroretinography (PERG), which measures the electrical activity of retinal ganglion cells, have demonstrated decreased PERG amplitudes in PD patients [21].

OCTA studies have shown reduced retinal vascular density in PD patients, particularly in the superficial capillary plexus [22]. This may reflect the microvascular dysfunction and reduced cerebral blood flow that are known to occur in PD. However, the exact relationship between retinal vascular changes and the pathophysiology of PD remains unclear. Longitudinal studies are needed to determine whether retinal vascular changes precede or follow the onset of motor symptoms in PD.

3.3. Multiple Sclerosis

Multiple sclerosis (MS) is a chronic inflammatory disease of the CNS characterized by demyelination and axonal damage. Optic neuritis, an inflammation of the optic nerve, is a common manifestation of MS. Retinal imaging, particularly OCT, has become an important tool for monitoring axonal damage in MS patients [23].

Several studies have shown that MS patients, particularly those with a history of optic neuritis, exhibit reduced RNFL thickness, as measured by OCT [11, 24]. RNFL thinning is thought to reflect axonal damage in the optic nerve and CNS. The degree of RNFL thinning correlates with the severity of MS and the degree of disability. OCT can also be used to monitor the progression of MS over time and to assess the response to treatment. Furthermore, ganglion cell layer thinning has been shown to be a feature of MS patients even in the absence of a prior history of optic neuritis suggesting retinal measurements may represent a valuable biomarker of subclinical disease progression [25].

OCTA studies have shown reduced retinal vascular density in MS patients, particularly in the peripapillary region [26]. This may reflect the microvascular damage that occurs in the CNS due to inflammation and demyelination. However, the exact relationship between retinal vascular changes and the pathophysiology of MS remains unclear.

Many thanks to our sponsor Esdebe who helped us prepare this research report.

4. Retinal Imaging in Systemic Vascular Diseases

The retinal microvasculature is a direct reflection of the systemic circulation, making retinal imaging a valuable tool for assessing cardiovascular risk and detecting early signs of systemic vascular diseases. Several studies have demonstrated the utility of retinal imaging in predicting cardiovascular events and monitoring the progression of systemic vascular diseases.

4.1. Diabetic Retinopathy

Diabetic retinopathy (DR) is a microvascular complication of diabetes that affects the blood vessels of the retina. DR is a leading cause of blindness worldwide. Retinal imaging, including fundus photography, FA, and OCTA, is essential for diagnosing and monitoring DR [2]. Fundus photography is used to detect microaneurysms, hemorrhages, exudates, and neovascularization. FA is used to visualize areas of capillary non-perfusion and leakage. OCTA is used to visualize and quantify microvascular changes, such as capillary dropout, microaneurysms, and neovascularization.

OCTA has emerged as a valuable tool for early detection of DR. It can detect subtle microvascular changes that may not be visible on fundus photography or FA. OCTA can also be used to assess the severity of DR and to monitor the response to treatment. Quantitative analysis of OCTA images, such as measuring the foveal avascular zone (FAZ) area and vascular density, can provide objective measures of DR progression.

4.2. Hypertension

Hypertension, or high blood pressure, can damage the blood vessels of the retina, leading to hypertensive retinopathy. Fundus photography is used to detect signs of hypertensive retinopathy, such as arteriolar narrowing, arteriovenous nicking, and hemorrhages [27].

OCTA has been used to assess retinal vascular density in hypertensive patients. Some studies have shown reduced retinal vascular density in hypertensive patients compared to healthy controls [28]. This may reflect the microvascular damage that occurs due to chronic hypertension. However, the relationship between retinal vascular changes and the severity of hypertension is still being investigated.

4.3. Cardiovascular Disease

Retinal imaging has been investigated as a potential tool for predicting cardiovascular events, such as heart attack and stroke. Several studies have shown that retinal vascular parameters, such as arteriolar and venular caliber, are associated with cardiovascular risk factors and cardiovascular events [29]. For example, narrower retinal arteriolar caliber has been associated with increased risk of hypertension, stroke, and coronary heart disease.

AI-based analysis of retinal images is being developed to automatically extract retinal vascular parameters and to predict cardiovascular risk. These AI-based systems have the potential to improve the accuracy and efficiency of cardiovascular risk assessment [30].

Many thanks to our sponsor Esdebe who helped us prepare this research report.

5. Challenges and Limitations

While retinal imaging offers numerous advantages as a diagnostic and monitoring tool, it also faces several challenges and limitations that need to be addressed. These challenges include image quality, patient variability, the need for standardized protocols, and the interpretation of results.

5.1. Image Quality

The quality of retinal images can be affected by various factors, including media opacities (e.g., cataracts), pupil size, and patient cooperation. Media opacities can scatter and absorb light, reducing the signal-to-noise ratio and making it difficult to visualize fine retinal structures. Small pupil size can limit the amount of light entering the eye, reducing image brightness and contrast. Patient movement and poor fixation can cause image blurring and artifacts.

Strategies to improve image quality include using mydriatic agents to dilate the pupils, improving patient positioning and fixation, and using advanced image processing techniques to reduce noise and artifacts. Furthermore, newer imaging technologies, such as swept-source OCT, offer better penetration through media opacities.

5.2. Patient Variability

Retinal structure and function can vary significantly among individuals due to factors such as age, ethnicity, refractive error, and coexisting ocular diseases. Age-related changes, such as RNFL thinning, can confound the interpretation of retinal imaging results. Ethnic differences in retinal pigmentation and vascular density can also affect image quality and quantitative measurements. Refractive error can affect the magnification of retinal images, requiring correction for accurate measurements. Coexisting ocular diseases, such as glaucoma and AMD, can alter retinal structure and function, making it difficult to isolate the effects of neurological or systemic diseases.

To address patient variability, it is important to collect detailed patient history and demographic information. Age-matched and ethnicity-matched control groups should be used for comparison. Refractive error should be corrected during image acquisition or analysis. The presence of coexisting ocular diseases should be carefully considered when interpreting retinal imaging results.

5.3. Standardized Protocols

Lack of standardized imaging protocols and analysis techniques can lead to variability in results and difficulty in comparing data across different studies. Different imaging devices and software platforms may use different algorithms and parameters for image acquisition and analysis. This can result in inconsistent measurements and unreliable conclusions.

To address this issue, it is important to develop and implement standardized imaging protocols and analysis techniques. This includes defining specific image acquisition parameters, such as scan size, resolution, and field of view. It also includes developing standardized algorithms for image analysis, such as segmentation and quantification. Multicenter studies and collaborative efforts are needed to validate and refine these standardized protocols.

5.4. Interpretation of Results

The interpretation of retinal imaging results can be challenging, particularly in the context of neurological and systemic diseases. Retinal changes may be subtle and non-specific, making it difficult to distinguish them from normal variations or coexisting ocular diseases. The relationship between retinal changes and disease progression may not be fully understood. Furthermore, the sensitivity and specificity of retinal imaging for detecting neurological and systemic diseases may be limited.

To improve the interpretation of retinal imaging results, it is important to correlate retinal findings with clinical data, such as patient history, physical examination, and other diagnostic tests. Longitudinal studies are needed to determine the natural history of retinal changes and their relationship to disease progression. Multimodal imaging approaches, combining different retinal imaging modalities and other imaging techniques, can provide a more comprehensive assessment of retinal structure and function. AI-based analysis of retinal images can help to identify subtle patterns and features that may be missed by human observers.

Many thanks to our sponsor Esdebe who helped us prepare this research report.

6. Future Directions

The field of retinal imaging is rapidly evolving, with new technologies and analysis techniques constantly being developed. Future directions in retinal imaging include the development of advanced imaging modalities, AI-driven analysis, and multimodal imaging approaches.

6.1. Advanced Imaging Modalities

Several advanced imaging modalities are being developed to improve the resolution, sensitivity, and specificity of retinal imaging. These include:

• High-resolution OCT: Advances in OCT technology are leading to higher resolution images, enabling visualization of finer retinal structures and earlier detection of subtle changes.
• Adaptive optics OCT (AO-OCT): AO-OCT combines the high resolution of adaptive optics with the cross-sectional imaging capabilities of OCT, enabling detailed visualization of retinal microstructure at the cellular level.
• OCT angiography (OCTA) with wider field of view: Expanding the field of view of OCTA allows for visualization of the peripheral retina, which is particularly useful in detecting peripheral vascular abnormalities.
• Functional retinal imaging: Techniques such as optophysiology and retinal metabolic imaging are being developed to assess retinal function and metabolism, providing insights into disease mechanisms and potential therapeutic targets.

6.2. AI-Driven Analysis

AI and machine learning algorithms are being used to automate image analysis, improve diagnostic accuracy, and predict disease progression. AI-based systems can automatically segment retinal layers, quantify vascular parameters, and detect subtle features that may be missed by human observers. AI can also be used to integrate retinal imaging data with other clinical data to predict disease risk and treatment response.

The use of deep learning algorithms has shown promising results in detecting DR, glaucoma, and AMD. These algorithms can be trained on large datasets of retinal images to identify patterns and features associated with these diseases. AI-based systems have the potential to improve the efficiency and accuracy of retinal image analysis, reducing the burden on clinicians and improving patient outcomes.

6.3. Multimodal Imaging Approaches

Combining different retinal imaging modalities and other imaging techniques can provide a more comprehensive assessment of retinal structure and function. Multimodal imaging approaches can help to overcome the limitations of individual imaging modalities and to improve diagnostic accuracy.

For example, combining OCT and OCTA can provide complementary information about retinal structure and vasculature. Combining retinal imaging with other imaging techniques, such as magnetic resonance imaging (MRI) and positron emission tomography (PET), can provide insights into the relationship between retinal changes and brain pathology. Multimodal imaging approaches have the potential to improve our understanding of disease mechanisms and to develop more effective diagnostic and therapeutic strategies.

Many thanks to our sponsor Esdebe who helped us prepare this research report.

7. Conclusion

Retinal imaging has emerged as a powerful tool for investigating systemic and neurological health. Its non-invasive nature, accessibility, and ability to visualize microvascular and neural structures make it a valuable modality for early detection, monitoring, and risk stratification of a wide range of diseases. Advancements in retinal imaging technologies and analysis techniques are constantly expanding its scope and sensitivity.

While challenges and limitations exist, future directions in the field, such as the development of advanced imaging modalities, AI-driven analysis, and multimodal imaging approaches, hold great promise for improving diagnostic accuracy and predictive power. Retinal imaging has the potential to revolutionize the management of systemic and neurological diseases, ultimately improving patient outcomes.

Many thanks to our sponsor Esdebe who helped us prepare this research report.

References

[1] Wilkinson, C. P., Ferris, F. L., 3rd, Klein, R. E., Lee, P. P., Agardh, C. D., Davis, M., … & Weber, M. L. (2003). Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. Ophthalmology, 110(9), 1677-1682.

[2] American Academy of Ophthalmology. (2023). Preferred Practice Pattern Guidelines. Diabetic Retinopathy. San Francisco, CA: American Academy of Ophthalmology.

[3] Huang, D., Swanson, E. A., Lin, C. P., Schuman, J. S., Stinson, W. G., Chang, W., … & Fujimoto, J. G. (1991). Optical coherence tomography. Science, 254(5035), 1178-1181.

[4] Spaide, R. F., Klancnik, J. M., Jr, & Cooney, M. J. (2015). Retinal vascular layers imaged by optical coherence tomography angiography. JAMA ophthalmology, 133(1), 45-50.

[5] Roorda, A., Penisten, D. K., Hofer, H., Williams, D. R., & Swanson, W. H. (2002). Adaptive optics scanning laser ophthalmoscopy. Vision research, 42(8), 1081-1092.

[6] Abràmoff, M. D., Garvin, M. K., Sonka, M., Niemeijer, M., Quellec, G., Vujosevic, S., … & Quellec, A. L. (2010). Retinal imaging and image analysis. IEEE reviews in biomedical engineering, 3, 169-208.

[7] Silva, P. S., Cavallerano, J. D., Sun, J. K., & Aiello, L. P. (2012). Peripheral lesions detected by ultrawide field imaging predict increased risk of diabetic retinopathy progression over 4 years. Ophthalmology, 119(7), 1272-1278.

[8] Yannuzzi, L. A., Rohrer, K. T., Tindel, L. J., Shapiro, H. M., Zuckerman, B. D., & Gilbert, J. L. (1985). Fluorescein angiography. Retina, 5(1), 1-22.

[9] Flower, R. W., Hochheimer, B. F., & Lutty, G. A. (1985). Indocyanine green dye fluorescence angiography. Retina, 5(1), 29-37.

[10] Wojtkowski, M., Leitgeb, R., Kowalczyk, A., Bajraszewicz, T., Radzewicz, C., & Fercher, A. F. (2002). In vivo human retinal imaging by Fourier domain optical coherence tomography. Journal of Biomedical Optics, 7(3), 457-463.

[11] Petzold, A., de Boer, J. F., Schippling, S., Vermersch, P., Kardon, R., Green, A. J., … & Calabresi, P. A. (2010). Optical coherence tomography in multiple sclerosis: a systematic review and meta-analysis. The Lancet Neurology, 9(9), 921-931.

[12] den Haan, J., Verbraak, F. D., Visser, P. J., Bouwman, F. H., van der Flier, W. M., Rozemuller, A. J., … & Scheltens, P. (2012). Retinal nerve fiber layer thickness in Alzheimer’s disease. Alzheimer’s & dementia, 8(3), 187-193.

[13] Selkoe, D. J. (2011). Alzheimer’s disease. Cold Spring Harbor perspectives in biology, 3(7), a004457.

[14] La Morgia, C., Barboni, P., Rizzo, G., Carbonelli, M., Regine, E., Caporali, L., … & Carelli, V. (2016). Retinal amyloid pathology in Alzheimer disease. Annals of neurology, 79(5), 717-727.

[15] Cheung, C. Y., Ong, S. J., Ikram, M. K., Gupta, P., Lee, T., & Tan, M. S. (2015). Retinal microvascular network and cognitive function: the Singapore Malay Eye Study. Journal of the American Heart Association, 4(4), e001592.

[16] Bayhan, H. A., Aslan Bayhan, S., Kurtuluş, F., Oztürk, B. T., & Gürdal, C. (2014). Ganglion cell–inner plexiform layer thickness in Alzheimer’s disease. Journal of ophthalmology, 2015.

[17] Bulut, M., Kurt, A., Sobaci, G., Mutlu, U., & Cakmak, H. B. (2018). Optical coherence tomography angiography findings in patients with Alzheimer disease. Retina, 38(2), 331-336.

[18] Dauer, W., & Przedborski, S. (2003). Parkinson’s disease. Neuron, 39(6), 889-909.

[19] Satue, M., Rodrigo, M. J., Obeso, J. A., & Marin, J. (2017). Optical coherence tomography as a potential biomarker for Parkinson’s disease. Parkinsonism & related disorders, 38, 53-58.

[20] Inzelberg, L., Ramirez, J. A., Haber, M., Asher, E., Jacob, G., Abboud, M., … & Chetrit, A. (2015). Retinal nerve fibre layer thinning in Parkinson disease. Vision research, 114, 253-259.

[21] Bodis-Wollner, I., Tagliati, M., Geller, A. M., Yahr, M. D., & Parkinson Study Group. (1999). Pattern electroretinography reveals reduced contrast sensitivity in Parkinson’s disease. Journal of the neurological sciences, 167(2), 127-134.

[22] Archibald, N. K., Clarke, J. R., Mosimann, U. P., Burn, D. J., & Lawden, M. C. (2020). Retinal vascular density in Parkinson’s disease using optical coherence tomography angiography. Journal of Parkinson’s Disease, 10(1), 291-298.

[23] Frohman, E. M., Dwyer, M. G., O’Connor, P. W., Buckle, G. J., Galetta, S. L., Calabresi, P. A., … & Balcer, L. J. (2006). Relationship of optic nerve and retinal measures to global brain atrophy in multiple sclerosis. Archives of neurology, 63(7), 1003-1009.

[24] Burkholder, B. M., Osborne, B., Loguidice, M. J., Frohman, E. M., Conger, A., Newsome, S. D., … & Balcer, L. J. (2009). Macular volume measurements by optical coherence tomography predict visual loss in multiple sclerosis. Archives of neurology, 66(11), 1361-1368.

[25] Saidha, S., Sotirchos, E. S., Oh, J., et al. Relationships between retinal axonal and ganglion cell layer damage and brain atrophy in multiple sclerosis. Neurology. 2011;77(23):2164-2172.

[26] Garcia-Martin, E., Larrosa, J. M., Polo, V., Herrero, R., Satue, M., Bambo, M. P., … & Pablo, L. E. (2016). Retinal and choroidal blood flow measured with optical coherence tomography angiography in multiple sclerosis. Investigative ophthalmology & visual science, 57(9), 5145-5152.

[27] Wong, T. Y., Klein, R., Couper, D. J., Cooper, L. S., Hubbard, L. D., & Samsa, G. P. (2004). Retinal microvascular abnormalities and incident stroke: the Atherosclerosis Risk in Communities Study. The Lancet, 363(9411), 938-944.

[28] Cheung, N., Liew, G., Lindley, R. I., & Wong, T. Y. (2010). Retinal microvascular signs and cardiovascular risk factors: the Blue Mountains Eye Study. American journal of ophthalmology, 149(5), 741-749.

[29] Ikram, M. K., de Jong, F. J., Vingerling, J. R., Hofman, A., de Jong, P. T., & Witteman, J. C. (2006). Retinal vessel diameters and risk of incident stroke: the Rotterdam Study. Stroke, 37(2), 489-494.

[30] Poplin, R., Varadarajan, A. V., Blumer, K., Bu, D., Baker, E., Mahmoudi, M., … & Peng, L. (2018). Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning. Nature biomedical engineering, 2(3), 158-164.