Methods for identifying cardiovascular disease from retinal scans

Abstract

It has been established that the presence of RIPLs is associated with higher prevalence of stroke, and especially in patients with atrial fibrillation. Enhanced methods for the detection and quantitation of retinal ischemic perivascular lesions (RIPLs) from commonly used OCT scans which are widely used in optometry and ophthalmology clinics are described. Systems and methods to characterize disease state and risk of stroke based on the identification and quantitation of RIPLs in the eyes of patients having atrial fibrillation have been developed. Methods of using information obtained according to the described methods to guide patient stratification for anticoagulation therapy are provided. Methods of identifying RIPLs from OCT scans are provided.

Description

[0001]METHODS FOR IDENTIFYING CARDIOVASCULAR DISEASE FROM RETINAL SCANS CROSS-REFERENCE TO RELATED APPLICATIONS This application claims benefit of and priority to U.S. Provisional Application No. 63 / 514,928, filed July 21, 2023, the content of which is hereby incorporated herein by reference in its entirety. STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH This invention was made with government support under P30CA016359 awarded by the National Institutes of Health; under K23 DK129836 awarded by the National Institute of Diabetes and Digestive and Kidney Diseases; and under R21EY035090 awarded by National Eye Institute. The government has certain rights in the invention. FIELD OF THE INVENTION The invention is generally related to systems and methods for early detection of cardiovascular diseases, particularly the detection of early-stage markers of cardiovascular disease by retinal scanning of patients having pre-clinical or sub-clinical indications of cardiovascular disease. BACKGROUND OF THE INVENTION Cardiovascular disease is the number one cause of death. Atrial fibrillation (AF) is the most common disorder of the heart rhythm in adults. The lifetime risk of development of AF is estimated to be about 1 in 4 for individuals above 40 years of age. Individuals with AF are at higher risk of developing complications from microemboli, such as strokes, heart failure and dementia. To reduce the risk of strokes and embolic events, anticoagulation therapy is often prescribed. While anticoagulation therapy reduces the risk of thromboembolism in individuals with AF, it is also associated with significant adverse events, most notably intracranial hemorrhage. Individuals with AF are often older in age, at higher risk of falls, and have significant co-morbidities such as hypertension and coronary artery disease. Consequently, they are at higher risk of complications from long-term anticoagulation therapy. In clinical practice, a commonly used tool to identify individuals with AF who are at greatest risk of developing strokes, and thus may benefit from anticoagulation, is the CHA2DS2-VASc score (Congestive heart failure, Hypertension, Age ≥ 75, Diabetes mellitus, Stroke or transient ischemic attack, Vascular disease, Age 65-74 years, Sex category). It is a point-based scoring system that incorporates the demographic information and clinical risk 45667975.1 1 factors, including age and sex, and presence of hypertension, diabetes, congestive heart failure, or prior stroke that are associated with risk stratification. Higher CHA2DS2-VASc scores are associated with higher risk of stroke, and anticoagulation is generally recommended for men and women with scores greater than 0 and 1, respectively. In addition to stroke, individuals with AF are also at risk of developing retinal vascular occlusions, and those with retinal artery occlusions are at higher risk of developing strokes. Despite this clinical tool, there is underutilization of anticoagulation in treating patients with AF, and approximately 40% of patients who are at high risk for stroke are not receiving anticoagulation. Although currently recommended by the guidelines, the CHA2DS2-VASc score has modest discrimination for ischemic stroke in patients with AF with a c-statistic of 0.675 in a meta-analysis evaluating eight clinical studies. As such, the identification of novel markers and the development of new screening tools are necessary to improve tailored anticoagulation strategies for the prevention of AF-related stroke. Patients with AF are also at risk of developing retinal vascular occlusions, and in a retrospective, cohort study, those with retinal artery occlusions had a higher incidence of new stroke compared to matched controls. Besides retinal artery or vein occlusions, there is also evidence of subclinical retinal ischemia in individuals with AF. The key in preventing or delaying the onset and development of cardiovascular disease is early detection and treatment. Unfortunately for many, the disease may go undiagnosed until the occurrence of serious events such as myocardial infarction or stroke. Identifying biomarkers of subclinical ischemia can help identify patients with occult cardiovascular disease. Previous studies have identified retinal ischemic perivascular lesions (RIPLs) as a retinal imaging biomarker of cardiovascular disease (Long, et al., EClinicalMedicine33 (2021) 100775; and Medala, et al., Ophthalmology Retina, V.6 (9), September 2022, Pages 865-867). RIPLs can be identified from optical coherence tomography (OCT) scans, which are widely used in ophthalmology and optometry clinics. While this discovery established the link between RIPLs and cardiovascular diseases, there are two significant challenges remaining: firstly, it is not clear how to easily identify RIPLs on optical coherence tomography (OCT) scans; secondly it is not clear how to extrapolate the clinical significance of RIPLs with various clinical conditions, especially in those with established cardiovascular disease. 45667975.1 2 Therefore, there is a need for improved systems and methods for the robust, reproducible, high-resolution detection and quantitation of RIPLs and for methods and systems to reliably and reproducibly transform this information to provide clinically- actionable information for individuals with cardiovascular disease, or who are at risk of developing cardiovascular disease, and / or who are at increased risk of developing significant complications related to cardiovascular diseases which include strokes, heart attacks, peripheral vascular disease, or retinal vascular occlusions. Thus, it is an object of the invention to provide systems and reagents for the improved detection, measurement and quantitation of retinal ischemic perivascular lesions. It is another object of the invention to provide improved systems and reagents to identify a subject as having or as being at increased risk of cardiovascular diseases such as coronary artery disease, atrial fibrillation, carotid disease, peripheral artery disease, or stroke, based on the characterization of retinal ischemic perivascular lesions in the subject. It is another object of the invention to provide improved systems and reagents to inform the status and / or monitor the progression of cardiovascular disease in a subject having pre-clinical or clinical cardiovascular diseases such as coronary artery disease, atrial fibrillation, carotid disease, peripheral artery disease, or stroke, based on the characterization of retinal ischemic perivascular lesions in the subject. It is yet another object of the invention to provide improved systems and methods to inform treatment regimens in subjects identified as having or being at increased risk of cardiovascular disease based on the characterization of retinal ischemic perivascular lesions in the subject. It is another object of the invention to provide enhanced systems and methods to guide patient stratification for therapy such as anticoagulant therapy and other therapies to optimize cardiovascular health, including statin therapy, cholesterol and lipid lowering drugs, surgical intervention for cardiovascular disease complications including heart rhythm control, coronary, carotid or peripheral vascular stents. BRIEF SUMMARY OF THE INVENTION It has been established that the presence of RIPLs is associated with higher prevalence of stroke. This association has also been shown in patients with atrial fibrillation. Enhanced methods for the detection and quantitation of retinal ischemic perivascular lesions (RIPLs) from commonly used OCT scans which are widely used in optometry and ophthalmology clinics are described. As demonstrated in the Examples, systems to characterize disease state 45667975.1 3 and risk of stroke based on the identification and quantitation of RIPLs in the eyes of patients having atrial fibrillation have been developed. Methods of using information obtained according to the described methods to guide patient stratification for anticoagulation therapy are provided. Methods of determining a treatment regimen for a subject are provided. The methods typically include the steps of (i) determining the presence or absence of retinal ischemic perivascular lesions (RIPLs) in the eye of the subject via macular optical coherence tomography (OCT); and optionally (ii) administering a therapy comprising anti-ischemia therapy and / or anti-stroke therapy to the subject when RIPLs are determined to be present in the eye of the subject. In some forms, the subject has, or has been identified as being at increased risk of having atrial fibrillation (AF), and / or other cardiovascular disease (CVD). In other forms, the subject does not have atrial fibrillation (AF) and / or other cardiovascular disease (CVD). In some forms, the methods include, prior to step (ii), determining additional risk stratification scores such as the CHA₂DS₂-VASc score, wherein the therapy is administered to the subject only when both (a) RIPLs are determined to be present in the eye of the subject and (b) the CHA₂DS₂-VASc score is 1 or above. In some forms, the subject is determined to have AF prior to step (ii). An exemplary therapy is anti-coagulation therapy. Methods for more effectively identifying a subject that has atrial fibrillation (AF) as being at risk of stroke are also described. In some forms, the methods include one or more steps of (i) determining the presence or absence of retinal ischemic perivascular lesions (RIPLs) in the eye of the subject via one or more systems for calculating the presence of RIPLs based on macular optical coherence tomography (OCT), and (ii) optionally determining a modified CHA₂DS₂-VASc score for the subject, wherein the modified CHA₂DS₂-VASc score excludes history of stroke, transient ischemic attack, and thromboembolism as risk factors, and wherein the modified CHA₂DS₂-VASc score includes a component based on the presence of RIPLs, determined in step (i). In some forms, the methods confirm that the subject is at increased risk of stroke when the modified CHA₂DS₂- VASc score is 1 or above. In some forms, the methods further include administering a therapy including anti-ischemia therapy and / or anti-stroke therapy to the subject based on the modified CHA₂DS₂-VASc score when RIPLs are determined to be present in the eye of the subject. Exemplary therapies include anti-platelet therapy, anti-ischemia therapy and / or anti- stroke therapy. In preferred forms, the therapy is anti-coagulation therapy. 45667975.1 4 In some forms, methods for more effectively identifying a subject that has atrial fibrillation (AF) as being at risk of stroke include the steps of: (i) determining the presence or absence of retinal ischemic perivascular lesions (RIPLs) in the eye of the subject, wherein the determination is made via macular optical coherence tomography (OCT) of an eye of the subject; and (ii) determining a risk stratification score such as CHA₂DS₂-VASc score. In some forms, the subject is identified as being at risk of stroke only when both (a) RIPLs are determined to be present in the eye of the subject and (b) the CHA₂DS₂-VASc score is 1 or above. In some forms, the methods further include administering a therapy comprising anti- ischemia therapy and / or anti-stroke therapy to the subject when the subject is identified as being at risk of stroke. An exemplary therapy is anti-coagulation therapy. In some forms, the methods can include (i) determining the presence or absence of retinal ischemic perivascular lesions (RIPLs) in an eye of a subject, wherein the subject has atrial fibrillation (AF); and (ii) selecting the subject for treatment for increased risk of stroke when the determining in step (i) identifies one or more RIPLs in an eye of the subject. In some forms, the methods can include, prior to step (ii), determining a modified “Congestive heart failure, Hypertension, Age≥75 years, Diabetes mellitus, Prior Stroke or TIA or Thromboembolism, Vascular disease, Age 65–74 years, Sex category” (CHA₂DS₂-VASc) score for the subject. In some forms, the methods can also include (iii) administering anti-ischemia treatment and / or anti-stroke treatment to the subject selected in step (ii) if the CHA₂DS₂- VASc score is 1 or above. In some forms, the treatment can include anti-coagulation therapy. In some forms, the anti-coagulation therapy can include administering an anticoagulant agent to the subject, wherein the anticoagulant agent is selected from the group consisting of warfarin, apixaban, dabigatran, edoxaban, rivaroxaban, heparin, and aspirin. In some forms, the treatment can include surgery, for example, to control heart rhythm. In some forms, for patients with or without atrial fibrillation the treatment can include cholesterol and lipid lowering drugs such as statin therapy. In some forms, the treatment can include surgery to place a stent in a major vessel in the heart, neck (carotid arteries) or peripheral vessels. In some forms, the methods can involve excluding the subject from a treatment or refraining from treating the subject when the subject is not selected in step (ii). In some forms, the methods can involve excluding the subject from a treatment or refraining from 45667975.1 5 treating the subject when the CHA₂DS₂-VASc score is below 1. In some forms, the method can involve discontinuing anti-ischemia treatment and / or anti-stroke treatment or refraining from further anti-ischemia treatment and / or anti-stroke treatment of the subject that is or had been receiving. In some forms, the determining in step (i) can include imaging one or both eyes of the subject. In some forms, the imaging can involve macular optical coherence tomography (OCT). In some forms, the OCT is one or more selected from spectral domain optical coherence tomography (SD-OCT), time domain optical coherence tomography (TD-OCT), swept source optical coherence tomography (SS-OCT), and visible light optical coherence tomography (VL-OCT). In some forms, the determining in step (i) can involve identification of RIPLs from data obtained by macular optical coherence tomography (OCT) and / or ultrasonography (USG) of an eye of the subject. In some forms, the determining in step (i) can involve detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) via visual inspection of images of the inner nuclear layer (INL), inner plexiform layer (IPL), outer nuclear layer (ONL) and / or outer plexiform layer (OPL) of an eye of the subject. In some forms, the visual inspection can involve duplicate, triplicate, or further inspections. In some forms, the duplicate, triplicate, or further inspections are carried out by two or more different inspectors. In some forms, the determining in step (i) involves detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) via visual inspection of images of the reconstructed en face image involving at least one of the inner nuclear layer (INL), inner plexiform layer (IPL), outer nuclear layer (ONL) and outer plexiform layer (OPL) of an eye of the subject. In some forms, the visual inspection of the en face image can involve duplicate, triplicate, or further inspections. In some forms, the duplicate, triplicate, or further inspections are carried out by two or more different inspectors. In some forms, the determining in step (i) includes (a) providing an OCT scan image of an eye of a subject; (b) identifying a cross section in the image corresponding to the retina which is flanked by the vitreous cavity and choroid; (c) fitting a polynomial curve to top points, bottom points and A center curve passing through the midpoint of the cross section to provide a fitted image; (d) optionally processing the fitted image to remove noise to provide a processed image, wherein the processing optionally includes one or more of thresholding, image dilation and erosion; (e) filtering the fitted and / or processed image to accentuate 45667975.1 6 horizontal lines using specified kernels to provide a filtered image, optionally wherein the filtering comprises Median filtering and / or Kernel filtering; (f) re-tracing lines and the center curve in the filtered image to provide a retraced image; and (g) identifying and / or quantifying RIPLs on the center lines and curve, optionally wherein the identifying includes normalization of slopes compared to reference line. In some forms, the method further includes one or more of (h) annotating the identified RIPLs on the original image, and / or quantitating and / or annotating the total number of RIPLs per scan; and / or (i) quantitating and / or recording a total number of RIPLs per eye and / or per subject. In some forms, the detecting in step (i) includes determining an RIPL index for a scan, wherein the RIPL index quantifies total positive deviation of interface between an outer plexiform layer (OPL) and an outer nuclear layer (ONL) above an average position for each B-scan. In some forms, determining the total positive deviation of the OPL / ONL interface above average includes (a) providing one or more OCT B-scan images of an eye of a subject, wherein a scan image includes a multiplicity of pixel intensities, wherein a pixel intensity of the multiplicity of pixel intensities includes a value, and wherein the multiplicity of pixel intensity values are disposed as columns and rows throughout the image; (b) processing the B-scans to reduce noise and highlight internal limiting membrane (ILM) and retinal pigment epithelium (RPE) layers in the images to provide a processed scan having columns of pixel intensity, optionally wherein the highlighting includes one or more processing steps of erosion, dilation, and / or thresholding; (c) segmenting the ILM and RPE, wherein the segmenting includes iterating down each column of pixel intensity values and up each column of pixel intensity values, respectively, until there is a change in pixel intensity value of at least a first threshold amount; (d) excluding scans of fovea, wherein the excluding comprises determining an average central retinal thickness and removing B-scans having a central retinal thickness that is a second threshold amount less than the average retinal thickness; (e) locating the darkest pixel in each column to identify the outer plexiform layer (OPL), wherein the darkest pixel has the greatest intensity value; (f) segmenting out the outer nuclear layer (ONL) / OPL interface, wherein the segmenting comprises iterating up each pixel in each column until a third threshold increase in pixel intensity; (g) calculating the pixel height between the RPE and ONL / OPL interface for each column; (h) optionally excluding heights several standard deviations above the mean to reduce errors in ONL / OPL interface detection; (i) obtaining positive deviations of the ONL / OPL interface, wherein the obtaining comprises determining the average height of the ONL / OPL interface above the 45667975.1 7 RPE, and subtracting the average height of the ONL / OPL interface above the RPE from the height at every column above the average; and (j) summing the positive deviations to provide the total volume of deviation of the OPL / ONL interface above average for each B-scan. In some forms of determining the total positive deviation of the OPL / ONL interface above average, step (b) further includes one or more step to filter out noise, optionally wherein the filtering includes a best fit spline having a single pixel thickness to trace the RPE; optionally further including tracing the OPL, wherein tracing the OPL includes increasing the thickness of the best fit spline from 1 pixel in thickness to up to about 7 pixels in thickness and moving the spline up by a proportion of the average distance between the RPE and ONL / OPL interface; optionally further including calculating the average pixel intensity for each column in the 7-pixel height RPE slice and 7-pixel height OPL slice, wherein columns where the RPE slice average pixel intensity are a fourth threshold value below average are excluded from the OPL slice to account for shadowing. In some forms, the method further includes counting multiple consecutive columns having pixel intensity above the fourth threshold value in the OPL slice as a single RIPL, and excluding multiple consecutive dark columns above a fifth threshold value, which is attributed to the fovea. In some forms, RIPLs are identified from an OCT scan utilizing a computer program run on a computer system, wherein the program is configured to identify RIPLs in the OCT scan by segmentation of the outer plexiform layer (OPL) which follows the curvature of the retinal pigment epithelium. In some forms, the method is repeated one or more times to sequentially identify RIPLs in scans from two or more subjects or pools of subjects. In some forms, the repeated method is carried out sequentially by a computer as a first and second or further iterations, wherein the computer implements a machine-learning system to increase the efficacy of the method in a second or further iteration as compared with a first or further iteration. In some forms, the method further includes, prior to step (ii), one or more steps to confirm that the subject has AF. In some forms, the determining in step (i) can involve detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) in the eye of the subject via a computer-based system for the automated analysis of images of the inner retinal layer (IRL) of an eye of the subject. In some forms, the computer-based system for the automated analysis of images of the inner retinal layer (IRL) can include (a) providing input data to a 45667975.1 8 computer processor operably connected to means for computing and displaying output data, and (b) computing the output data. In some forms, the input data can include one or more images of the inner retinal layer (IRL) of an eye of the subject, or data derived therefrom, where the output data comprises the number of RIPLs detected within the one or more images. In some forms, the input data can involve a computer-readable media. In some forms, the methods can include, prior to step (ii), one or more steps to confirm that the subject has AF. In some forms, the AF is asymptomatic AF or silent AF. In some forms, the subject can have a personal or family history of any one or more of AF, cardiovascular disease, stroke, or ischemic disease. In some forms, the subject can have an implantable cardiac device, and / or has had recent cryptogenic stroke or transient ischemic attack (TIA).In some forms, the subject has a personal or family history of diseases or disorder selected from the group consisting of retinopathy, diabetic retinopathy, hypertensive retinopathy, ischemic retinopathy, or macular edema. BRIEF DESCRIPTION OF THE DRAWINGS Figures 1A-1G are images showing Retinal Ischemic Perivascular Lesions (RIPLs) on spectral-domain (SD) OCT scans of study subjects. Fig.1A shows Infrared image of retina of the left eye of case 1 with no appreciable abnormalities. Fig.1B shows an SD-OCT B-scan of the same individual as in Fig.1A with a RIPL depicted by an arrowhead. Figs.1C- 1F show SD-OCT B-scans of different subjects (cases), including cases 3 (Fig.1C), 5 (Fig. 1D), 4 (Fig.1E), and 11 (Fig.1F), with Retinal ischemic perivascular lesions (RIPLs) denoted with arrowheads in each image, respectively. Cases 3, 5, 4, 11 represent individuals with and without atrial fibrillation, and with documented and undocumented cardiovascular disease, respectively. Fig.1G shows an en face 3D reconstruction, at the level of the outer plexiform layer demonstrating 3 RIPLs appearing as dark spots depicted with three arrowheads, with a perivascular predilection. Figures 2A-2B show a images obtained by Spectral domain optical coherence tomography (SD-OCT) B-scans, demonstrating Paracentral Acute Middle Maculopathy (PAMM) (Fig.2A), and Retinal Ishchemic Perivascular Lesions (RIPLs) (Fig.2B), respectively, with each of the inner nuclear layer, outer plexiform layer and outer nuclear layer indicated. As depicted in Fig.2A, two distinct PAMM lesions are indicated by arrows, characterized by increased hyperreflectivity within the inner nuclear layer. Progression to RIPLs is depicted in Fig.2B, as increased hyperreflectivity within the outer nuclear layer, indicated by two arrows. 45667975.1 9 Figure 3 is a Receiver operating characteristic (ROC) graph, showing sensitivity (0-1) over specificity. Curves for CHAD2S-VASC score, and, RIPLs + CHAD2S-VASC are indicated / labelled, together with a reference line, respectively. Figures 4A-4M are images of the same OCT scan, annotated to show one method for the detection of RIPLS. The original image (Figs.4A, 4C, 4E) is saved and cropped, and then annotated with best-fit curves for each of three regions of interest (Fig.4B), then further-processed from the original to de-noise the image, for example, to create a threshold image (Fig.4D), and an eroded image (Fig.4F), and a dilated image (Fig.4G), respectively. The de-noised image is then filtered in one or more steps to accentuate the horizontal lines from the original (Fig.4H), using Median filtering (Fig.4I) and Kernel filtering (Fig.4J), respectively. The lines in the kernel filtered image are traced and the line closer to the center line (identified in Fig.4B) is identified (Fig.4K), then RIPLs are identified as a “bump” in the horizontal line (Fig.4L) and quantified. For comparison, the identified RIPLs can be superimposed on the original, pre-processed image (Fig.4M). DETAILED DESCRIPTION OF THE INVENTION I. Definitions The term “sample” from a subject means a tissue (e.g., tissue biopsy), organ, cells (including a cell maintained in culture), cell lysate (or lysate fraction), or body fluid from a subject. Non-limiting examples of body fluids include blood, urine, plasma, serum, tears, lymph, bile, cerebrospinal fluid, interstitial fluid, aqueous or vitreous humor, colostrum, sputum, amniotic fluid, saliva, anal and vaginal secretions, perspiration, semen, transudate, exudate, and synovial fluid. In preferred embodiments, the biological sample of the disclosed methods is urine, or serum obtained from the subject, and lysates of bone marrow or peripheral blood cells. The term “subject” means any individual who is the target of diagnosis or treatment administration. The subject can be a vertebrate, for example, a mammal. Thus, the subject can be a human. The term does not denote a particular age or sex. A patient refers to a subject afflicted with a disease or disorder. The term “patient” includes human and veterinary subjects. The term “therapeutically effective” means that the amount of the composition used is of sufficient quantity to ameliorate one or more causes or symptoms of a disease or disorder. Such amelioration only requires a reduction or alteration, not necessarily elimination. 45667975.1 10 By “treat” or “treatment” is meant the medical management of a patient with the intent to cure, ameliorate, stabilize, or prevent a disease, pathological condition, or disorder. This term includes active treatment, that is, treatment directed specifically toward the improvement of a disease, pathological condition, or disorder, and also includes causal treatment, that is, treatment directed toward removal of the cause of the associated disease, pathological condition, or disorder. In addition, this term includes palliative treatment, that is, treatment designed for the relief of symptoms rather than the curing of the disease, pathological condition, or disorder; preventative treatment, that is, treatment directed to minimizing or partially or completely inhibiting the development of the associated disease, pathological condition, or disorder; and supportive treatment, that is, treatment employed to supplement another specific therapy directed toward the improvement of the associated disease, pathological condition, or disorder. “Inhibit,” “inhibiting,” and “inhibition” mean to decrease an activity, response, condition, disease, or other biological parameter. This can include but is not limited to the complete ablation of the activity, response, condition, or disease. This may also include, for example, a 10% reduction in the activity, response, condition, or disease as compared to the native or control level. Thus, the reduction can be a 10, 20, 30, 40, 50, 60, 70, 80, 90, 100%, or any amount of reduction in between as compared to native or control levels. By “pharmaceutically acceptable” is meant a material that is not biologically or otherwise undesirable, i.e., the material may be administered to a subject, along with the nucleic acid or vector, without causing any undesirable biological effects or interacting in a deleterious manner with any of the other components of the pharmaceutical composition in which it is contained. The term “En face image” as used herein means that the reconstruction of individual B-scans from OCT scans. These images can be reconstructed from software available through commercial OCT devices or through third party softwares. Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein. Use of the term “about” is intended to describe values either above or below the stated value in a range of approx. + / - 10%; in other embodiments the values may range in value either above or below the stated value in a range of approx. + / - 5%; in other embodiments the 45667975.1 11 values may range in value either above or below the stated value in a range of approx. + / - 2%; in other embodiments the values may range in value either above or below the stated value in a range of approx. + / - 1%. The preceding ranges are intended to be made clear by context, and no further limitation is implied. The use of any and all examples, or exemplary language (e.g., “such as”) provided herein, is intended merely to better illuminate the description and does not pose a limitation on the scope of the description unless otherwise claimed. Disclosed are materials, compositions, and components that can be used for, can be used in conjunction with, can be used in preparation for, or are products of the disclosed method and compositions. These and other materials are disclosed herein, and it is understood that when combinations, subsets, interactions, groups, etc. of these materials are disclosed that while specific reference of each various individual and collective combinations and permutation of these compounds may not be explicitly disclosed, each is specifically contemplated and described herein. For example, if a ligand is disclosed and discussed and a number of modifications that can be made to a number of molecules including the ligand are discussed, each and every combination and permutation of ligand and the modifications that are possible are specifically contemplated unless specifically indicated to the contrary. Thus, if a class of molecules A, B, and C are disclosed as well as a class of molecules D, E, and F and an example of a combination molecule, A-D is disclosed, then even if each is not individually recited, each is individually and collectively contemplated. Thus, in this example, each of the combinations A-E, A-F, B-D, B-E, B-F, C-D, C-E, and C-F are specifically contemplated and should be considered disclosed from disclosure of A, B, and C; D, E, and F; and the example combination A-D. Likewise, any subset or combination of these is also specifically contemplated and disclosed. Thus, for example, the sub-group of A-E, B- F, and C-E are specifically contemplated and should be considered disclosed from disclosure of A, B, and C; D, E, and F; and the example combination A-D. Further, each of the materials, compositions, components, etc. contemplated and disclosed as above can also be specifically and independently included or excluded from any group, subgroup, list, set, etc., of such materials. These concepts apply to all aspects of this application including, but not limited to, steps in methods of making and using the disclosed compositions. Thus, if there are a variety of additional steps that can be performed it is understood that each of these additional steps can be performed with any specific embodiment or combination of embodiments of the 45667975.1 12 disclosed methods, and that each such combination is specifically contemplated and should be considered disclosed. All methods described herein can be performed in any suitable order unless otherwise indicated or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g., “such as”) provided herein, is intended merely to better illuminate the embodiments and does not pose a limitation on the scope of the embodiments unless otherwise claimed. No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the invention. II. Improved Methods for detection of Cardiovascular Disease Methods for the rapid and effective detection of subclinical cardiovascular diseases and disorders in a subject based on inspection of an eye of the subject have been established. In some forms, the subject is a subject with subclinical indications of cardiovascular diseases and disorders who has no diagnosis of cardiovascular disease (CVDs) and may be unaware that they have cardiovascular diseases or disorders. In some forms, the subject is a patient with one or more clinical indication(s) of cardiovascular disease, such as atrial fibrillation (AF) and who has no other diagnosis of cardiovascular disease (CVDs) and whom is unaware that they have another cardiovascular disease or disorder. In other forms, the subject is a patient with clinical indication of cardiovascular disease, such as atrial fibrillation (AF) and who is at increased risk of an ischemic disease, such as a TIA or a stroke. Therefore, methods for the identification of occult cardiovascular disease (CVDs) in subjects are described. In some forms, the methods inform the presence of CVD in the subject to facilitate early institution of medical interventions and lifestyle modifications. The methods are designed based on the discovery that individuals with subclinical cardiovascular disease are at higher risk of developing retinal artery and vein occlusions, and even in the absence of overt retinal vascular occlusions, these patients may exhibit signs of subclinical retinal ischemia. Therefore, the methods identify and quantify subclinical retinal ischemia in subjects to identify subjects having, or at increased risk of having CVD. In some forms, the methods include one or more steps of: (i) selecting a subject; (ii) investigating the retina of one or more eyes of the subject for the presence of one or more Retinal Ischemic Perivascular Lesions (RIPLs) in the eye of the subject; and (iii) identifying the subject as having, or as being at increased risk of cardiovascular disease (CVD) and / or ischemic disease based on the observation of one or more RIPLs in the eye of the subject. 45667975.1 13 In some forms, the methods include one or more additional steps of: (iv) performing one or more additional methods on the subject to identify or confirm the subject as having, or being at increased risk of having CVD; and (v) administering one or more active agents or procedures to the subject to treat or prevent CVD in the subject. The methods can include any technique known in the art for the detection of RIPLs in the eye of a subject. For example, in some forms, the methods are applied to detect RIPLs using data obtained from a standard or routine techniques for imaging an eye of a subject. In some forms, the methods employ one or more systems configured to correlate the incidence of RIPLs with one or more clinical indications in the subject. In some forms, the methods include calculating a modified form of the CHA2DS2-VASc risk assessment score for a subject. For example, in some forms, the methods calculate a modified risk assessment score that includes all of the “classical” data that is incorporated into a conventional CHA2DS2-VASc, as well as data for the presence and number of RIPLs identified within one or both eyes of a subject. In some forms, inclusion of data for incidence of RIPLs with OCT improves discrimination of stroke in individuals with AF beyond the conventional CHA2DS2- VASc score. In some forms, the modified CHA2DS2-VASc score is used to identify a subject as significantly associated with increased risk of stroke or TIA. Therefore, in some forms, the methods include generating a modified CHA2DS2-VASc score to identify a subject requiring prophylactic treatment for stroke and / or TIA. In some forms, when the methods include administering a treatment to the subject, the treatment includes anticoagulant therapy. Each of the method steps is described in more detail, below. A. Selecting a Subject Any of the described methods can include the step of selecting a subject. In some forms, the step of selecting a subject includes selection based on the presence or absence of one or more factors, such as a prior medical diagnosis, familial history, age, sex, geographic location, or a lifestyle factor, such as smoking, consumption of alcohol, high blood pressure, thyroid disease or other chronic health conditions, obesity, etc. Typically, a subject in need of the detection and quantitation of subclinical retinal ischemia is a subject who has been diagnosed as having or identified as being at increased risk of having a cardiovascular disease (CVD). In other forms, the subject is a normal, healthy control subject who has been 45667975.1 14 identified as not having CVD. In other forms, the subject is a patient in need of a routine or investigative examination of one or more eyes, regardless of CVD status. 1. Subjects having or at increased risk of Cardiovascular Disease (CVD) In some forms, the subject has previously been diagnosed as having, or identified as being at increased risk of having a cardiovascular disease (CVD). In addition to atrial fibrillation, there are four main types of CVD, including coronary heart disease, stroke, peripheral arterial disease and aortic disease. In some forms, the subject has previously been diagnosed as having, or identified as being at increased risk of having coronary heart disease (CHD). CHD occurs when the heart muscle’s blood supply is blocked or interrupted by a build-up of fatty substances (atheroma) in the coronary arteries. The coronary arteries are the major blood vessels that supply the heart with blood. If the coronary arteries become narrow, e.g., due to a build-up of atheroma, the blood supply to the heart muscle is restricted, leading to angina (chest pains). If a coronary artery becomes completely blocked, it can cause a heart attack. In some forms, the subject has previously been diagnosed as having, or identified as being at increased risk of having a stroke. A stroke occurs when the blood supply to the brain is disturbed, preventing the constant supply of oxygen and nutrients that are required for correct brain function, leading to the death and damage of brain cells. This can lead to brain damage and possibly death. The sooner a subject having a stroke receives treatment, the less damage is likely to occur. The main stroke symptoms include bodily and / or facial paralysis; inability to move or raise the legs or arms due to weakness or numbness; slurred or garbled speech or inability to talk. In some forms, the subject has previously been diagnosed as having, or identified as being at increased risk of having peripheral arterial disease, also known as peripheral vascular disease. Peripheral arterial disease occurs when there is a blockage in the arteries to the limbs (typically legs). The most common symptom of peripheral arterial disease includes pain in the legs when walking, including cramp, a dull pain or a sensation of heaviness in the muscles of the legs. In some forms, the subject has previously been diagnosed as having, or identified as being at increased risk of having Aortic disease. The aorta is the largest blood vessel in the body, which carries blood from the heart to the rest of the body. The most common type of 45667975.1 15 aortic disease is aortic aneurysm, which is where the wall of the aorta becomes weakened and bulges outwards. Typical symptoms include pain in the chest, back or abdomen. In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having atrial fibrillation (AF). In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having a transient ischemic attack (TIA). In some forms, the subject has previously been diagnosed as having, or previously identified as being at increased risk of having a CVD-associated indication, such as a heart attack, but has not been diagnosed with AF or stroke. In some forms, the subject has previously been diagnosed as having, or previously identified as being at increased risk of having a CVD-associated indication, such as a heart attack, but has not been diagnosed with AF or stroke. In other forms, the subject has a sub-clinical indication of CVD, for example, which does not manifest in physical symptoms that are typically identified by one or more other analytical methodologies. For example, in some forms, the subject has not previously been diagnosed as having, or identified as being at increased risk of having a CVD-associated indication. For example, in some forms, the subject has not previously been diagnosed as having, or previously identified as being at increased risk of having atrial fibrillation (AF). In some forms, the subject has not previously been identified as having, or identified as being at increased risk of having a stroke. In some forms, the subject has previously been diagnosed as having, or previously identified as being at increased risk of having a CVD-associated indication, such as a heart attack, but has not been diagnosed with AF or stroke. In some forms, the subject has previously been diagnosed as having, or previously identified as being at increased risk of having a CVD-associated indication, such as a heart attack, but has not been diagnosed with AF or stroke. In some forms, the subject has not previously been identified as having, or identified as being at increased risk of having a stroke, but has been diagnosed with carotid artery disease, or has been diagnosed with carotid artery disease including carotid artery stenosis. In some forms, the subject is receiving or has previously received treatment for a CVD-associated indication. An exemplary treatment is anti-coagulant therapy. In some forms, the subject has, or is suspected as having AF, CVD or stroke. In some forms, the subject is receiving or has previously received treatment for one or more of AF, CVD or stroke. 45667975.1 16 In some forms, the subject is an individual aged 50 to 90 years diagnosed with AF who has undergone optical coherence tomography (OCT). i. Atrial Fibrillation (AF) In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having atrial fibrillation (AF). Atrial fibrillation occurs when abnormal electrical impulses suddenly start firing in the atria (upper chambers of the heart). These impulses override the heart’s natural pacemaker, which can no longer control the rhythm of the heart. The atria contract randomly and sometimes so fast that the heart muscle cannot relax properly between contractions; the normal range for a heart rate is 60 to 100 beats a minute, whereas the heart rate in atrial fibrillation may range from 100 to 175 beats a minute. This reduces the heart's efficiency and performance and causes a highly irregular pulse rate. Because not enough blood is being pumped out of the atria, blood pools in the area. The pooled blood can clot which can lead to ischemia, for example, if a blood clot is pumped out of the heart to the brain, blocking the blood supply to the brain and causing a stroke. About 15% to 20% of people who have strokes have heart arrhythmia, and the risk of blood clotting is why patients with this condition are prescribed blood thinners. Untreated AF doubles the risk of heart-related deaths and is associated with a fivefold increased risk for stroke. The mean interval from initial diagnosis of atrial fibrillation to death is approximately 25.2±9.5 years. In some forms, the subject has been diagnosed with AF, or as being at increased risk of AF relative to a normal, healthy control subject. Exemplary symptoms of AF include sensations of a fast, fluttering or pounding heartbeat (palpitations); chest pain; dizziness; fatigue; lightheadedness; reduced ability to exercise; shortness of breath; and weakness. Therefore, in some forms, the subject has one or more symptoms selected from sensations of a fast, fluttering or pounding heartbeat (palpitations); chest pain; dizziness; fatigue; lightheadedness; reduced ability to exercise; shortness of breath; and weakness. In some forms, the subject has occasional symptoms of AF. In some forms, the subject has persistent symptoms of AF. In some forms, the subject has long-standing, persistent symptoms of AF, where one or more of the symptoms of AF persist for at least 6 months or at least 12 months. In some forms, the subject has permanent symptoms of AF. In some forms, the subject has AF but has not previously been diagnosed with AF or made aware that they may have AF. Therefore, in some forms, the subject has “silent” AF, which does not cause pain or concern to the subject and has not been noticed or diagnosed by 45667975.1 17 a clinician. Silent atrial fibrillation (AF) is an asymptomatic atrial arrhythmia. In some forms, the silent AF has some minor non-specific symptoms, such dizziness, a fluttering in the chest, repeated fatigue. Typically, AF and silent AF can be diagnosed by a clinician, for example, by a systematic electrocardiogram, an external Holter, or from implanted cardiac devices. In some forms, the subject has AF and / or silent AF that are difficult to diagnose. Therefore, in some forms, the subject suspects a diagnosis of AF or silent AF but is in need of one or more confirmatory tests. ii. Coronary Artery Disease In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having Coronary artery disease. Coronary artery disease is a common heart condition. The major blood vessels that supply the heart (coronary arteries) struggle to send enough blood, oxygen and nutrients to the heart muscle. Cholesterol deposits (plaques) in the heart arteries and inflammation are usually the cause of coronary artery disease. Signs and symptoms of coronary artery disease occur when the heart doesn’t get enough oxygen-rich blood. In coronary artery disease, reduced blood flow to the heart can cause chest pain (angina) and shortness of breath. A complete blockage of blood flow can cause a heart attack. Coronary artery disease often develops over decades. Symptoms may go unnoticed until a significant blockage causes problems or a heart attack occurs. iii. Carotid Artery Disease In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having carotid artery disease. Carotid artery disease occurs when fatty deposits, i.e., atherosclerotic plaques, reduce patency of the carotid arteries, blood vessels that deliver blood to the brain and head. Plaques increases the risk of stroke. Carotid artery disease often develops slowly. The first sign of the condition might be a stroke or transient ischemic attack (TIA). A TIA is a temporary shortage of blood flow to the brain. iv. Peripheral Arterial Disease In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having peripheral artery disease (also called peripheral arterial disease). Peripheral artery disease is a common condition in which narrowed arteries reduce blood flow to the arms or legs. In peripheral artery disease (PAD), one or more limbs (legs and / or arms) receive insufficient blood flow to keep up with demand. This may cause 45667975.1 18 leg pain when walking (claudication) and other symptoms. Peripheral artery disease is usually a sign of a buildup of fatty deposits in the arteries (atherosclerosis). Atherosclerosis causes narrowing of the arteries that can reduce blood flow in the legs and, sometimes, the arms. Typically, peripheral artery disease treatment includes exercising, eating a healthy diet and not smoking or using tobacco. v. Transient ischemic attack (TIA) In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having a transient ischemic attack (TIA). A transient ischemic attack (TIA) is a stroke that lasts only a few minutes. It occurs when the blood supply to part of the brain is briefly interrupted. TIA symptoms, which usually occur suddenly, are similar to those of stroke but do not last as long. Most symptoms of a TIA disappear within an hour, although they may persist for up to 24 hours. Symptoms of TIA include one or more of: numbness or weakness in the face, arm, or leg, especially on one side of the body; trouble seeing in one or both eyes; difficulty with walking, dizziness; confusion or difficulty in talking or understanding speech; and loss of balance and / or coordination. Therefore, in some forms, the subject has one or more symptoms of numbness or weakness in the face, arm, or leg, especially on one side of the body; trouble seeing in one or both eyes; difficulty with walking, dizziness; confusion or difficulty in talking or understanding speech; and loss of balance and / or coordination. vi. Ischemic Stroke In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having a stroke, such as a thromboembolic stroke. A stroke occurs when thrombi cause embolization to the cerebral circulation. Patients with atrial fibrillation (AF) have an increase in the incidence of stroke with 90% of thrombi (blood clots) originating in the left atrial appendage (LAA). The LAA is a tubular out-pouching of the left atrial chamber that has complex and highly variable anatomy, that is prone to blood stasis and subsequent blood clotting due to its narrow and long tubular connection with the atrium. In patients with atrial fibrillation, 91% of left atrial clots originate in the LAA. If these clots embolize to the cerebral vasculature, they have the potential to cause a stroke. Typical symptoms of ischemic stroke include sudden numbness or weakness of the face, arm or leg; impairment of sight in one or both eyes; impairment of speech; extreme headache; loss of balance; ataxia and / or impaired mobility. 45667975.1 19 In some forms, the subject has experienced a cryptogenic stroke. The term cryptogenic stroke, as used herein, designates the category of ischemic stroke for which no probable cause is found despite a thorough diagnostic evaluation. According to the American Stroke Association, some possible causes of cryptogenic stroke include: patent foramen ovale (e.g., a hole that develops between the left and right upper chambers of the heart); atrial fibrillation; and a fatty buildup of plaque inside the aorta. vii. Other Diseases In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having another disease that is a known complication associated with cardiovascular disease, such as cognitive impairment or dementia. Heart diseases include conditions that impairs cardiac function regardless of the specific modality that is affected. Heart disease is a growing problem, and the resultant vascular insufficiency has the potential to impair function in other organs, including the brain. There is an established body of literature implicating heart disease as a risk factor for dementia. In addition, a number of studies suggest that cardiovascular risk factors are independently associated with the development of dementia. These risk factors include hypertension, hypercholesterolemia, diabetes, obesity, and smoking. There are many types of dementias and a wide spectrum of vascular cognitive impairment, including Alzheimer’s disease (AD), vascular dementia (VaD), frontotemporal dementia, dementia with Lewy body, Parkinson’s disease-related dementia. Cognitive impairment is a broad term that generally describes a decline in cognitive functions, and the severity of the impairment ranges from mild cognitive impairment (MCI) to dementia, as well as the most common variety, mixed dementia that is characterized by multiple cerebral pathologies with prominent vascular involvement. Therefore, in some forms, the subject has been identified as having, or as being at increased risk of having cognitive impairment associated with one or more cardiovascular risk factors. Ranging from mild cognitive impairment (MCI) to dementia. In other forms, the subject has a personal or family history of diseases or disorder selected from the group consisting of any one or more of retinopathy, diabetic retinopathy, hypertensive retinopathy, ischemic retinopathy, or macular edema. viii. Implantable Cardiac Device In some forms, the subject has been previously diagnosed as having, or previously identified as being at increased risk of having CVD, and has an implantable cardiac device. 45667975.1 20 Exemplary cardiac implantable electronic devices include pacemakers, implantable cardioverter defibrillator (ICD), biventricular pacemakers, and cardiac loop recorders, or any other prostheses designed to help control or monitor irregular heartbeats in subjects with certain heart rhythm disorders and heart failure. Therefore, in some forms, the subject has a pacemaker, implantable cardioverter defibrillator (ICD), biventricular pacemaker, or cardiac loop recorder. In other forms, the subject has had cardiovascular surgery, for example, to install a stent or other patency device. B. Detection of Retinal Ischemic Perivascular Lesions (RIPLs) The methods include one or more steps of detecting Retinal Ischemic Perivascular Lesions (RIPLs) in one or both eyes of the subject. Retinal ischemic perivascular lesions, or RIPLs, represent focal atrophy of the middle retina or inner nuclear layer (INL). These lesions are a legacy of paracentral acute middle maculopathy (PAMM), which is thought to occur because of the preferential susceptibility of the inner nuclear layer, at the level of the deep capillary plexus, to ischemia, or cotton wool spots. Retinal hypoperfusion may occur secondary to a decrease in afferent blood flow into the retina capillary plexus in patients with various disorders such as low ejection fraction, carotid artery stenosis, thrombi or emboli formation, or vascular stasis, all of which have been associated with PAMM. Therefore, the eye manifests the earliest evidence of ischemia due to systemic disease at the level of the middle retina which is exquisitely sensitive to subtle changes in blood flow. While both PAMM and RIPLs can be readily visualized in a non-invasive manner, using spectral domain optical coherence tomography (SD-OCT) scans, only RIPLs are permanent. Hyperreflective PAMM lesions are transient and usually resolve within a six-week period but leave a legacy of INL thinning, with a compensatory expansion of outer plexiform layer or Henle’s fiber layer. Therefore, RIPLs represent an actionable imaging biomarker that can be harnessed to detect ischemia in the retina. 1. Scan Types Data obtained using any technique known in the art for investigating the structure of the eye of a subject can be used for the detection and / or quantitation of RIPLs according to the described methods. In preferred forms, the methods include assessing data obtained by one or more contemporary scanning techniques that are used in the clinic for visualizing and / or imaging the structure of the eye. Exemplary techniques include Optical coherence tomography (SD-OCT) scans, including but not limited to time-domain, spectral domain or swept source. 45667975.1 21 In some forms, the methods include detection of RIPLs in data obtained from an investigation of one or both eyes of the subject up to 10 years prior to the detection, for example, in some forms, the data were obtained (e.g., by OCT scanning of the eye of the subject up to one minute, one hour, one day, two days, three days, four days, five days, six days, one week, two weeks, three weeks, four weeks, one month, two months, three months, six months, nine months or one year, or 10 years prior to the detection. In some forms, the methods include one or more steps of obtaining data by scanning one or both or the eyes of the subject. i. Optical coherence tomography (OCT) scans In some forms, the methods detect RIPLs in data obtained using Optical coherence tomography (OCT) scanning of an eye of the subject. Optical coherence tomography is a method of using low-coherence interferometry to determine the echo time delay and magnitude of backscattered light reflected off an object of interest. This method can be used to scan through the layers of a structured tissue sample such as the retina with very high axial resolution (3 to 15 μm), providing images demonstrating 3D structure. Because of the unique optically clear pathway through the eye, OCT has been used most extensively for imaging disorders affecting the retina. It has also been used in the anterior segment of the eye as well as to view the structure of non-ocular tissues, including breast, kidney, and skin lesions, among others. In some forms, data from Time-domain Optical coherence tomography (TD-OCT) is used to detect RIPLs in the eye of a subject. TD-OCT is frequently compared to ultrasound because their basic principles are analogous, with OCT using light as its medium, whereas ultrasound uses sound. Both methods create a cross-sectional image by measuring the echo time delay and intensity of the reflected and backscattered light or sound. OCT images use this information to depict variations in optical reflectance through the depth of the tissue along a point, creating what is known as an A-scan. These single axial scans through the tissue can be gathered linearly across the tissue, making one cross-sectional image, known as a B-scan, and a collection of parallel B-scans can be used to gather a 3D data set. Because of the difference in medium, OCT has a much higher axial resolution than ultrasound, ~10 μm for TD-OCT1,and ~5 to 7 μm for SD-OCT7–10 vs ~150 μm for ultrasound at a frequency of 10 MHz. Higher axial resolutions can be achieved with higher frequencies of ultrasound. Ultrasound at 50 to 80 MHz has axial resolutions of ~50 to 20 μm, but with a penetration depth of only 4 to 3 mm. The use of light as the medium in OCT gives it the advantage of 45667975.1 22 being noncontact for the patient, whereas ultrasound needs a coupling medium such as water along its path between transmitter and tissue to pass the sound waves. These physical differences make ultrasound useful for measuring intraocular distances and overall structure of the anterior eye, with greater penetration of signal into the angle where near-infrared light is blocked by the sclera, but OCT is far more useful for detecting detailed structures in the retina and anterior segment. While the general principle of using reflections to create the images is the same for OCT and ultrasound, the methods for detecting these reflections are not. Because light is so much faster than sound, the time delays between reflections from different layers cannot be measured directly since differences would be on the order of femtoseconds. Optical coherence tomography uses low-coherence interferometry to see the time difference corresponding to the distances between structures. The process starts with a broad-bandwidth laser or superluminescent diode low-coherence light source, the beam from which travels to a beam splitter. One half of the light goes to a mirror at a known position on a reference arm, and the other goes to the sample arm, where it is scattered and reflects off of tissue structures. Light from the reference and sample arms travels back to the beam splitter and recombines to form an interference pattern, which is sensed by a photodetector. The light beams combine constructively only if the light from the tissue and the light from the reference mirror are at almost exactly the same distance. The width of the signal envelope defines the resolution of the interferometer and is determined by the coherence length of the light used; the shorter the coherence length, the finer the resolution. The coherence length is dependent on the bandwidth, with broader bandwidths producing lower-coherence light. In TD-OCT, for each axial scan, the position of the reference mirror is oscillated to allow imaging of a known range of depths in the tissue. Each pixel in the A-scan represents the intensity of the reflection at that position. This intensity is converted to a log scale because of the approximately 45-dB-wide variation in intensity. In some forms, data from Spectral-domain Optical coherence tomography (SD-OCT) is used to detect RIPLs in the eye of a subject. Spectral domain OCT (SD-OCT) increases axial resolution 2- to 3-fold and scan speed 60- to 110-fold vs time domain OCT (TD-OCT). SD-OCT enables novel scanning, denser sampling, and 3-dimensional imaging. OCT is a ubiquitous tool with fast image acquisition. The detection of RIPLs with OCT is a simple process that can have ramifications beyond the eye clinic, since RIPLs detection provides an 45667975.1 23 entry point for general medical practice in the management of cardiovascular health and can select subjects who would benefit from additional and targeted diagnostic testing. Exemplary OCT techniques include time domain OCT (TD-OCT) and spectral domain OCT (SD-OCT), which increases axial resolution 2- to 3-fold and scan speed 60- to 110-fold vs time domain OCT (TD-OCT). SD-OCT enables novel scanning, denser sampling, and 3-dimensional imaging. This thesis tests my hypothesis that SD-OCT improves reproducibility, sensitivity, and specificity for glaucoma detection. Therefore, in some forms, the methods detect RIPLs in data obtained by TD-OCT and / or SD-OCT. a. Swept-source (SS) OCT In some forms, the methods detect RIPLs in data obtained using Swept-Source Optical coherence tomography (OCT) scanning of an eye of the subject. Swept-Source is another OCT technique that can be used to obtain high-resolution images of the retina. Swept-Source OCT (SS-OCT) differs from other imaging technology in terms of the particles used. To produce images, MRI and ultrasound use sound waves whereas SS-OCT uses light waves. This is the prime difference between these imaging technologies. SS OCT produces a broad bandwidth in an optical setup that helps in obtaining the optical reflections required for imaging the structures. A narrow-band light source with a wavelength equivalent to 1000 nm (1 nanometer = 0.0000000001 m) is used in the swept- source OCT. Due to SS-OCT’s insensitivity roll-off, it allows simultaneous imaging of multiple structures in the eye of a subject. b. Visible-light OCT In some forms, the methods detect RIPLs in data obtained using Visible-light Optical coherence tomography (OCT) scanning of an eye of the subject. Visible-light OCT (vis-OCT) is another OCT technique that can be used to obtain high- resolution images of the retina. It relies on visible light illumination, whereas most commercial and investigational OCTs use near-infrared light. As a result, vis-OCT requires different considerations in engineering design and implementation but brings unique potential benefits to both fundamental research and clinical care of several diseases. Here, we intend to provide a summary of the development of vis-OCT and its demonstrated applications. We also provide perspectives on future technology improvement and applications. 2. Characterization of RIPLs from scan data The methods identify RIPLs from images showing the interior retinal surface and / or subretinal structures of the eye of a subject. 45667975.1 24 In some forms, identification of RIPLs is carried out by visual analysis of one or more OCT images showing the retinal layers in B-scans of OCTs of the eye of a subject. In other forms, the method is carried out by visual analysis of processed OCT images creating en face reconstruction of a layer containing any of the inner nuclear layer (INL), inner plexiform layer (IPL), outer plexiform layer (OPL) or outer nuclear layer (ONL), to identify RIPLs as hyporeflective loci from en face images data. In other forms, the methods employ one or more systems, such as automated or computational systems, to identify RIPLs from data, such as scan images. As demonstrated in the Examples, in exemplary forms, SD-OCT scans were acquired on a Spectralis SD-OCT machine (Heidelberg Engineering, Heidelberg, Germany) as part of a routine retinal examination of an eye of the subject. In some forms, RIPLs are identified using the previously published criteria (see, for example, Long, et al. Prevalence of subclinical retinal ischemia in patients with cardiovascular disease – a hypothesis driven study, EclinicalMedicine 2632021;33:100775, the content of which is incorporated herein by reference in its entirety). Typically, the image includes a cross sectional image of the interior retinal surface of the eye, for example, including structures that can be identified and discriminated as the inner nuclear layer, the outer plexiform layer and the outer nuclear layer of the eye. In an exemplary method, individual b-scans are reviewed by one, two, three, or more than three independent and masked observers. An RIPL is typically defined as the presence of focal atrophy or thinning of the inner nuclear layer with a compensatory expansion of the outer nuclear layer. In some forms, when more than one observer is involved, a consensus between all of the observers is required to identify an RIPL. Exemplary images, showing the presence of RIPLs as determined by visual inspection, is depicted in Figs.1B-1G and Fig.2. i. Exemplary Methods In some forms, RIPLs are identified and quantified in an eye of a subject using one or more of the 5 exemplary methods set forth below. Each of these methods is described as follows: a. Method 1 In a first method, RIPLs are identified from an OCT scan according to one or more of the following nine steps (i)-(ix), including: (i) Providing an OCT scan image of an eye of a subject. Typically an image (in any format) saved and cropped to exclude any additional labels. 45667975.1 25 (ii) Identifying the cross section in the image corresponding to the retina which is flanked by the vitreous cavity and choroid. Typically, this is achieved by identifying peaks in the vertical lines. An exemplary image is depicted in Fig.4B, showing the best fit curves (top and bottom) and the center line of a cropped image (Fig 4A). Typically, the cross section includes structures that can be identified and discriminated as the inner nuclear layer, the outer plexiform layer and the outer nuclear layer of the eye. (iii) Fitting a polynomial curve to the top points, the bottom points and the center curve passing through the midpoint of the window to provide a fitted image. Typically, the location of the center curve is defined to be present exactly between the top and bottom best fit line or up to 75% closer to the top or bottom line depending on the eye anatomy. (iv) Optionally, additional processing to remove noise from the fitted image to provide a processed image. In some forms, method optionally includes additional pre- processing steps to remove “noise” from the image (i.e., denoising). Exemplary de-noising steps include one or more of image thresholding, image dilation and erosion. One or more of these steps typically removes noise from the image (see, e.g., Figs 4C-4G). (v) Filtering the fitted and / or processed image to accentuate horizontal lines using specified kernels (see, e.g., Figs.4H-4J) to provide a filtered image. Exemplary additional filtering that can be applied to denoise the image and allow identification of horizontal lines includes Median filtering and Kernel filtering. (vi) Re-tracing lines and the center curve in the filtered image to provide a retraced image. In an exemplary method, the center lines and curve are identified in step (iii) (see, e.g., Fig.4K). (vii) Identifying and / or quantifying RIPLs as a “bump” on the center lines and curve. The measurement of bumps and assignment of an RIPL can be carries out using different methods. In some forms, the identifying includes normalization of slopes compared to reference line. (viii) In some forms, the methods optionally include annotating the identified RIPLs on the original image, and / or quantitating and / or annotating the total number of RIPLs per scan. (ix) In some forms, the methods further include quantitating and / or recording a total number of RIPLs per eye and / or per subject to provide a dataset for one or both eyes for one or more subjects. In some forms, the data are recorded and provided in the form of a database. 45667975.1 26 b. Method 2 In a second method, RIPLs are identified from an OCT scan utilizing a computer vision heuristic model to calculate an “RIPL index” for each OCT volume (i.e., for each B- scan). An RIPL index serves as a measure of the total positive deviation of the ONL / OPL interface above its average position. For each B-scan, erosion, dilation, and thresholding are used to reduce noise and highlight pertinent layers. The internal limiting membrane (ILM; the basal lamina of the inner retina) and Retinal pigment epithelium (RPE; formed from a single layer of regular polygonal cells arranged at the outermost layer of the retina) are then segmented out by iterating down each column and up each column, respectively until there is a significant threshold change in pixel intensity. B-scans with a central retinal thickness that is a certain amount less than the average retinal thickness are excluded, as a method for excluding scans of the fovea. The darkest pixel is then located in each column to identify the OPL. The ONL / OPL interface is then segmented out by iterating up each pixel in each column until a threshold increase in pixel intensity. The pixel height between the RPE and ONL / OPL interface is then calculated for each column. Errors in ONL / OPL interface detection are reduced by excluding heights several standard deviations above the mean. The average height of the ONL / OPL interface above the RPE is subtracted from the height at every column to obtain the deviation of the ONL / OPL interface at every column above the average. Positive deviations are then summed to get the total volume of deviation of the OPL / ONL interface above average for each B-scan. This value is then averaged for all 49 B- scans in an OCT acquisition to calculate the RIPL index. Therefore, the second method includes: (i) providing one or more OCT B-scans of an eye of a subject; (ii) processing each B-scan to reduce noise and highlight pertinent layers, for example using erosion, dilation, and thresholding; (iii) segmenting the ILM and RPE out by iterating down each column and up each column, respectively until there is a significant threshold change in pixel intensity; (iv) excluding scans of the fovea by removing B-scans with a central retinal thickness that is a certain amount less than the average retinal thickness; (v) locating the darkest pixel in each column to identify the OPL; (vi) iterating up each pixel in each column until a threshold increase in pixel intensity to segment out the ONL / OPL interface; 45667975.1 27 (vii) calculating the pixel height between the RPE and ONL / OPL interface for each column; (viii) optionally excluding heights several standard deviations above the mean to reduce errors in ONL / OPL interface detection; (ix) Subtracting the average height of the ONL / OPL interface above the RPE from the height at every column to obtain the deviation of the ONL / OPL interface at every column above the average; and (x) summing the positive deviations to provide the total volume of deviation of the OPL / ONL interface above average for each B-scan. The total volume of deviation value is then averaged for all 49 B-scans in an OCT acquisition to calculate the RIPL index. c. Method 3 In a third method, RIPLs are identified from an OCT scan utilizing a computer vision heuristic model to count the number of RIPLs in each OCT volume (B-scans). The third method incorporates the same steps (i) to (ix) as in method 2, above, to segment out the RPE and the OPL / ONL interface. A best fit spline is then used to better trace the RPE and filter out noise. The best fit spline is then increased from 1 pixel in thickness to 7 pixels in thickness. The slice of 7 pixels in height is then moved up by a proportion of the average distance between the RPE and ONL / OPL interface to trace the OPL. The average pixel intensity for each column in the 7- pixel height RPE slice and 7-pixel height OPL slice are then calculated. Columns where the RPE slice average pixel intensity are a certain proportion below average are excluded from the OPL slice to account for shadowing. Multiple consecutive dark columns in the OPL slice are counted as a single RIPL, excluding multiple consecutive dark columns above a certain threshold which is attributed to the fovea. d. Method 4 In a fourth method, RIPLs are identified from an OCT scan utilizing a machine learning object detection model to detect RIPLs, as set forth in the Examples. In an exemplary form, multiple B-scans from OCT scan of a multiplicity of eyes from a multiplicity of subjects are preprocessed to crop the image to focus on the layers of interest using part of the computer vision algorithm described in Method 2, above. A 30-pixel height slice that traced the RPE is shifted up a certain amount to capture the ONL, OPL, and INL. That portion of the image is cropped. The cropped image is then 45667975.1 28 sliced in three parts horizontally and then stacked vertically to create an image closer to the dimensions of a square which could then be fed into the object detection model. Labeled bounding boxes are moved accordingly with the preprocessing. In an exemplary method, an SSD MobileNet V2 FPNLite 320x320 COCO17 TPU-8 pretrained model is used as the starting point for the model. The SSD MobileNet V2 FPNLite is a state-of-the-art object detection model that combines the strengths of the single-shot multibox detector (SSD) and MobileNet V2, a lightweight and efficient convolutional neural network (CNN) architecture. This model is particularly suitable for the task, due to its fast inference time and high accuracy, which are crucial for processing large volumes of OCT images in a clinical setting. It is based on the MobileNet V2 architecture, which employs depth-wise separable convolutions and an inverted residual structure, enabling it to achieve high performance with a relatively low number of parameters. This reduces the computational complexity of the model, making it ideal for deployment on edge devices and real-time applications. In addition, the model incorporates a feature pyramid network (FPN), which is a multi-scale feature extraction technique that enhances the detection of objects at different scales and aspect ratios. This is particularly beneficial for the task, as RIPLs can vary in size and shape, making them challenging to detect consistently. By integrating the FPN with the SSD detection framework, the model is able to identify RIPLs more effectively and with improved precision. The model is trained on a local machine having suitable computing power and suitable software. An exemplary local machine is a computer running Ubuntu 20.04 with an NVIDIA T4004GB GPU using Python version 3.8, Tensorflow 2.11.1, CUDA version 12.0, and CUDNN version 8.8.1. An exemplary batch size is 32. In some forms, the model is trained over 10 epochs; typically a low number of epochs is chosen to avoid overfitting because the images with RIPLs had been duplicated to even the distribution between non- RIPL and RIPL containing images. e. Method 5 In a fifth method, RIPLs are identified from an OCT scan utilizing a segmentation of the outer plexiform layer which follows the curvature of the retinal pigment epithelium (or alternatively choriocapillaris or internal limiting membrane). Following the curvature allows to avoid sudden deviations from RIPLs. Alternative methods construct a best fit line at the level of the outer plexiform layer. The area is increased 45667975.1 29 to thickness ranging from 1 pixel to 30 pixels, or to a thickness in between. RIPLs are identified as dark spots in the line containing the outer plexiform layer. C. Identifying the subject as having or at risk of CVD The methods include one or more steps to correlate the detection of RIPLs from images obtained from OCT scans showing the interior retinal surface and / or subretinal structures of the eye of a subject without known disease, or with one or more disease states of the subject. In some forms, the methods correlate the presence and / or number of RIPLs identified in one or both eyes with a risk of CVD in a subject having no prior knowledge or diagnosis of CVD risk. In other forms, the methods correlate the presence of and / or number of RIPLs identified in one or both eyes with a diagnosis of CVD in a subject having prior knowledge or a diagnosis of CVD or CVD risk. In some forms, the methods correlate the presence and / or number of RIPLs identified in one or both eyes in a subject having no prior knowledge or diagnosis of CVD risk factors (such as hypertension, diabetes, hypercholesterolemia, hyperlipidemia). In other forms, the methods correlate the presence of and / or number of RIPLs identified in one or both eyes with a diagnosis of CVD risk factors (such as hypertension, diabetes, hypercholesterolemia, hyperlipidemia) in a subject having prior knowledge or a diagnosis of CVD risk factors (such as hypertension, diabetes, hypercholesterolemia, hyperlipidemia). In other forms, the methods correlate the presence of and / or number of RIPLs identified in one or both eyes with a diagnosis of CVD risk factors (such as hypertension, diabetes, hypercholesterolemia, hyperlipidemia) in a subject not having prior knowledge or a diagnosis of CVD risk factors (such as hypertension, diabetes, hypercholesterolemia, hyperlipidemia). 1. Input Data In some forms, the methods correlate the presence and / or number of RIPLs identified in one or both eyes with CVD, or a risk of CVD in a subject, based on two or more pieces of input data relating to the subject. The input data include: (i) the number of RIPLs identified in a scan of one eye of the subject, for example, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more than 10 RIPLs; and (ii) optionally one or more factors associated with the subject, including prior medical diagnosis, familial history, age, sex, geographic location, or a lifestyle factor, such as smoking, consumption of alcohol, high blood pressure, hypertension, diabetes, 45667975.1 30 hypercholesterolemia, hyperlipidemia, thyroid disease or other chronic health conditions, obesity, etc. In some forms, the methods employ one or more systems or algorithms to provide a correlation between diagnosis of CVD, type of CVD, and / or risk of CVD in the subject. Typically, the input data is device-agnostic. For example, the methods can include data obtained on macular scans of the middle retinal layer with various parameters and using scans exported from a variety of different commercially available devices. The methods associate RIPL data together with additional data relating to the subject, such as the presence of atrial fibrillation and / or stroke to inform additional clinical utility in the management of the subject with arrythmia who may benefit from anticoagulation. In some forms, the presence or absence of RIPLs is correlated with the absence or presence of CVD. In some forms, the presence or absence of RIPLs is correlated with the absence or presence of CVD risk factors. In other forms, the presence or absence of RIPLs is indicative of an increased risk for CVD complications, such as stroke, heart attack, complications from peripheral artery disease, and / or complications from carotid artery stenosis. In other forms, presence of RIPLs in both eyes as opposed to one eye (right or left) is indicative of the presence of CVD. Typically, CVD risk factors are indicative of increased risk for CVD complications such as stroke, heart attack, complications from peripheral artery disease, and / or complications from carotid artery stenosis. In other forms, an increased number of RIPLs (such as 2, 3, 4, 5, 6, 7, 8, 9, 10 or more than 10) in one or both eyes (right or left) of a subject is indicative of the presence of CVD. Typically, CVD risk factors indicate increased risk for CVD complications such as stroke, heart attack, and / or complications from peripheral artery disease, and / or complications from carotid artery stenosis. 2. Detecting diseases and disorders The methods analyze and / or interpreting the scan data and conclude one or more findings about, e.g., the health status of the subject. Therefore, the disclosed methods include the determination, identification, indication, correlation, diagnosis, prognosis, etc. (which can be referred to collectively as “identifications”) of subjects, diseases, conditions, states, etc. based on imaging, measurements, detections, comparisons, analyses, assays, screenings, etc., of RIPLs in the eye(s) of the subject and optionally one or more additional input data points relating to the subject. 45667975.1 31 Therefore, in some embodiments, the methods include one or more steps of concluding about the health status of the subject. In some forms, the disclosed methods identify subjects having as being at increased risk of AF. In some forms, the disclosed methods identify subjects having AF as being at increased risk of ischemic disease or stroke. Such identifications are useful for many reasons. For example, and in particular, such identifications allow specific actions to be taken based on, and relevant to, the particular identification made. For example, diagnosis of a particular disease or condition, such as AF and / or risk of stroke in particular subjects (and the lack of diagnosis of that disease or condition in other subjects) has the very useful effect of identifying subjects that would benefit from treatment, actions, behaviors, etc. based on the diagnosis. For example, treatment for a particular disease or condition in subjects identified is significantly different from treatment of all subjects without making such an identification (or without regard to the identification). Subjects needing or that could benefit from the treatment will receive it and subjects that do not need or would not benefit from the treatment will not receive it. Accordingly, also disclosed are methods including taking particular actions following and based on the disclosed identifications. For example, disclosed are methods including creating a record of an identification (in physical—such as paper, electronic, or other—form, for example). Thus, for example, creating a record of an identification based on the disclosed methods differs physically and tangibly from merely performing an imaging, measurement, detection, comparison, analysis, assay, screen, etc. Such a record is particularly substantial and significant in that it allows the identification to be fixed in a tangible form that can be, for example, communicated to others (such as those who could treat, monitor, follow-up, advise, etc. the subject based on the identification); retained for later use or review; used as data to assess sets of subjects, treatment efficacy, accuracy of identifications based on different measurements, detections, comparisons, analyses, assays, screenings, etc., and the like. For example, such uses of records of identifications can be made, for example, by the same individual or entity as, by a different individual or entity than, or a combination of the same individual or entity as and a different individual or entity than, the individual or entity that made the record of the identification. The disclosed methods of creating a record can be combined with any one or more other methods disclosed herein, and in particular, with any one or more steps of the disclosed methods of identification. As another example, disclosed are methods including making one or more further identifications based on one or more other identifications. For example, particular treatments, 45667975.1 32 monitoring, follow-ups, advice, etc. can be identified based on the other identification. For example, identification of a subject as having a disease or condition with a high level of a particular component or characteristic can be further identified as a subject that could or should be treated with a therapy based on or directed to the high-level component or characteristic. A record of such further identifications can be created (as described above, for example) and can be used in any suitable way. Such further identifications can be based, for example, directly on the other identifications, a record of such other identifications, or a combination. Such further identifications can be made, for example, by the same individual or entity as, by a different individual or entity than, or a combination of the same individual or entity as and a different individual or entity than, the individual or entity that made the other identifications. The disclosed methods of making a further identification can be combined with any one or more other methods disclosed herein, and in particular, with any one or more steps of the disclosed methods of identification. In some forms, the methods identify a subject as having subclinical CVD. In other forms, the methods identify a subject as having clinical CVD, for example, in some forms a subject that has previously been diagnosed as having subclinical CVD is confirmed as having subclinical CVD, or as having developed CVD. In some forms, the methods identify a subject as having silent CVD, such as silent AF. In some forms, the methods confirm a subject as having silent CVD, such as silent AF, for example, where the same subject has previously been identified as being at risk of CVD and / or AF. In some forms, the methods identify a subject having silent CVD, such as silent AF, as being at increased risk of ischemic disease, such as thromboembolic stroke. Generally, the methods identify a subject as being at “increased risk” of developing a disease or disorder when the subject is identified as being at higher likelihood of developing the disease or disorder than a control subject. In some forms, control subjects include normal, healthy individuals. In other forms, the control subject is not healthy, for example, in some forms, the control has silent AF, or silent CVD, or AF, or CVD, but does not have the same phenotype or genotype or other characteristics or conditions that increase the likelihood of further or different disease or disorders in the subject. In some forms, when the methods identify one or more RIPLs in the eye of a subject with established AF, the methods identify the subject as being at increased risk of stroke, for example, as compared with a control subject having established AF and having fewer or no RIPLs in one or both eyes. 45667975.1 33 Exemplary diseases associated with RIPLs include cardiovascular diseases (coronary artery disease, AF, stroke, peripheral artery disease, carotid disease, stroke, TIA, heart arrythmia), as well as non-cardiovascular diseases, including any disease or conditions that lead to or associated with decreased blood flow to the retina. Diseases and disorders associated with decreased blood flow to the retina include, diabetes mellitus, diabetes mellitus type 1, diabetes mellitus type 2, hypertension, hypercholesterolemia, diabetic retinopathy, nonproliferative diabetic retinopathy (mild, moderate or severe), proliferative diabetic retinopathy, any retinopathy, uveitis, macular edema, retinal vein occlusion (branch or central), retinal artery occlusion (branch or central), paracentral acute middle maculopathy, glaucoma. Therefore, in some forms, the methods identify a subject with or without CVD as having, or as being at increased risk of having one or more of diabetes mellitus, diabetes mellitus type 1, diabetes mellitus type 2, hypertension, hypercholesterolemia, diabetic retinopathy, non-proliferative diabetic retinopathy (mild, moderate or severe), proliferative diabetic retinopathy, any retinopathy, uveitis, macular edema, retinal vein occlusion (branch or central), retinal artery occlusion (branch or central), paracentral acute middle maculopathy, and glaucoma. 3. Risk score In some forms, the methods include one or more steps of determining a risk score, such as a Modified or unmodified CHA₂DS₂-VASc score for a subject. Typically, the Modified CHA₂DS₂-VASc score is based on the CHA₂DS₂-VASc score (See, e.g., Friberg, et al., European Heart Journal, Volume 33, Issue 12, June 2012, Pages 1500–1510, the content of which is incorporated herein by reference in its entirety). i. CHA2DS2-VASc Score The CHADS2score and its updated version, the CHA2DS2-VASc score, are clinical prediction rules for estimating the risk of stroke in people with AF. C = Congestive heart failure; H = Hypertension: blood pressure consistently above 140 / 90 mmHg (or treated hypertension on medication); A = Age ≥75 years; D= Diabetes mellitus; S2 = Prior Stroke or TIA or Thromboembolism; V = Vascular disease (e.g. peripheral artery disease, myocardial infarction, aortic plaque); A = Age 65–74 years; Sc = Sex category (i.e. female sex). Typically, a CHA₂DS₂-VASc score is used to determine whether or not treatment is required with anticoagulation therapy or antiplatelet therapy, since AF can cause stasis of blood in the upper heart chambers, leading to the formation of a mural thrombus that can dislodge into the blood flow, reach the brain, cut off supply to the brain, and cause a stroke. A 45667975.1 34 high score corresponds to a greater risk of stroke, while a low score corresponds to a lower risk of stroke. An exemplary stroke risk evaluation based on multiple criteria is set forth in Table 1, below. Table 1: Annual stroke risk (%) according to CHA2DS2-VASc score CHA2DS2-VASc score Friberg 2012 Lip 2010 95% CI 0 0.2 0.0 0.0–0.0 1 0.6 0.6 0.0–3.4 2 2.2 1.6 0.3–4.7 3 3.2 3.9 1.7–7.6 4 4.8 1.9 0.5–4.9 5 7.2 3.2 0.7–9.0 6 9.7 3.6 0.4–12.3 7 11.2 8.0 1.0–26.0 8 10.8 11.1 0.3–48.3 9 12.2 100 2.5–100 Typically, the CHA₂DS₂-VASc score is used to determine whether oral anticoagulant, with well controlled vitamin K antagonist (VKA, e.g. warfarin with time in therapeutic range >70%), or a direct oral anticoagulant (DOAC, e.g. dabigatran, rivaroxaban, edoxaban or apixaban). An exemplary indication for therapy based on the CHA₂DS₂-VASc score is set forth in Table 2. Table 2: CHA₂DS₂-VASc score and Therapy determination Score Risk Therapy Indicated 0 (male) or 1 (female) Low No anticoagulant therapy 1 (male) Moderate Oral anticoagulant should be used 2 or greater High Oral anticoagulant recommended a. Modified CHA2DS2-VASc Score including RIPL Data (CHA2D-VASc) In some forms, the methods include steps to provide a risk assessment and / or inform optimal therapeutic regimens in a subject. In some forms, the methods provide a risk assessment and / or inform optimal therapeutic regimens in a subject by providing a modified CHA2DS2-VASc Score for the subject to account for additional data associated with detection of one or more RIPLs in one or both eyes of the subject. 45667975.1 35 As described in the Examples, in some forms, a modified CHA2DS2-VASc Score is calculated, using a multivariable logistic regression to evaluate the relationship of RIPLs with stroke. In some forms, the methods calculate the modified CHA2DS2-VASc Score in which history of stroke, transient ischemic attack or thromboembolism may or may not be excluded as a risk factor, and where RIPLs may or may not be incorporated. In an exemplary form, a modified CHA2DS2-VASc Score is termed a CHA2D-VASc, in which history of stroke, transient ischemic attack or thromboembolism are excluded as a risk factor. For example, in some forms, multifarious nature of data prohibit the determining of timing of strokes. Therefore, in some forms, a multivariable logistic regression is used to evaluate the relationship of RIPLs with stroke, after adjustment for CHA2D-VASc score. As demonstrated in the Examples, the area under the curve (AUC) is calculated for two models: CHA2D-VASc score alone and CHA2D-VASc score + RIPLs (see, Figure 3), and these models were compared using the Likelihood Ratio Test. In this model, each additional CHA2D-VASc point was associated with OR 1.40 (1.03, 1.91) of stroke, and presence of RIPLs was associated with OR 2.87 (1.26, 6.55) of stroke (p=0.03, 0.01 respectively, see Table 8). The modified score, which incorporated both CHA2D-VASc and RIPLs, was superior to the model utilizing modified CHA2D-VASc alone based on the Likelihood Ratio test (p=0.01). Accordingly, when combining presence of RIPLs with CHA2D-VASc in a receiver operating characteristic analysis, or ROC analysis, the area under the curve (AUC) in determining stroke was 0.70. This AUC was an increase from 0.62 when using CHA2D-VASc alone (as indicated in Figure 3). As used herein, a receiver operating characteristic ROC analysis involves creating a ROC curve, i.e., a graphical plot that illustrates the performance of a binary classifier model at varying threshold values. The ROC curve is the plot of the true positive rate against the false positive rate at each threshold setting. ROC analysis provides a systematic tool for quantifying the impact of variability among individuals' decision thresholds. In some forms, creating a modified CHA2DS2-VASc Score includes calculating the area under the curve (AUC) for a CHA2D-VASc Score + RIPLs. In some forms, creating a modified CHA2D-VASc Score includes addition of 0, 1, 2, or 3 points according to presence or absence of RIPLs, in one eye, or both eyes, above a certain threshold which can be 0, 1, 2, 3, 4, 5, 6, 7, 8 , 9, 10, or more than 10 RIPLs. Typically, both RIPLs and high CHA2DS2- VASc Score are significantly associated with stroke. 45667975.1 36 Therefore, in some forms, the methods determined a modified risk score for a subject that includes a CHA2D-VASc in addition to one or more pieces of datum associated with presence of one or more RIPL in an eye of the subject. In some forms, the presence of one or more RIPL in an eye of the subject increases likelihood of stroke or TIA or thromboembolic events in the subject by an odds ratio (OR) value. In some forms, the presence of one or more RIPL in an eye of the subject increases likelihood of stroke or TIA or other thromboembolic events in the subject by an odds ratio (OR) of between about 1.1 and about 7.0. For example in some forms, the presence of one or more RIPL in an eye of the subject increases likelihood of stroke or TIA or other thromboembolic events in the subject by an odds ratio (OR) of between the 1.26 and about 6.55. In an exemplary form, the presence of one or more RIPL in an eye of the subject increases likelihood of stroke or TIA or other thromboembolic events in the subject by an odds ratio (OR) of 2.87. Typically, the calculation of an Odds Ratio (OR) is carried out according to the equation: OR= (odds of the event in the exposed group) / (odds of the event in the non- exposed group). As described in the Examples, each additional CHA2DS2-VASc Score point is associated with Odds Ratio (OR) 1.40 (1.03, 1.91) of stroke, and presence of RIPLs was associated with OR 2.87 (1.26, 6.55) of stroke, respectively. Thus using the odds ratio, this group of subjects having RIPLs has 2.87 times the odds of having a stroke than those lacking RIPLs. In some forms, the methods identify the presence of subclinical retinal ischemia in a subject with atrial fibrillation. In some forms, the methods identify the subject as having an increased risk of stroke. In some forms, the methods identify the subject as needing anticoagulation therapy, and / or as being a subject for whom the benefit from anticoagulant therapy will outweigh the undesired side effects. D. Treatment Regimens In some forms, the methods include one or more steps of determining a treatment regimen for a subject, based on the presence of RIPLs in one or both eyes of the subject. In some forms, the methods incorporate a modified CHA2DS2-VASc Score to assist in the determination of a treatment regimen for a subject. In some forms, the presence of one, or more than one, such as two, three, four, five, six, or more than six RIPLs in one or both eyes of the subject is associated with the need for treatment. In some forms, the presence of no RIPLs or a single RIPL or more RIPLs in one or both eyes of a subject excludes the subject from one or more treatment regimens. 45667975.1 37 1. Treating a Subject In some forms, the methods include one or more steps of treating a subject identified as having or as being at increased risk of having AF or a CVD. Therefore, methods of treatment for AF, silent AF,CVD or silent CVD are provided. In some forms, the methods include treating a subject for AF, silent AF, CVD or silent CVD based on presence of one, or more than one, such as two, three, four, five, six, or more than six RIPLs in one or both eyes of the subject. In other forms, the presence of no RIPLs or a single RIPL in one or both eyes of a subject excludes the subject from one or more treatment regimens. The methods typically treat a subject by administering one or more therapeutic agents and / or procedures to the subject. In exemplary forms, the methods administer one or more reagents or regimens designed to prevent ischemic disease and thromboembolic stroke in a subject identified as having one or more RIPLs in one or both eyes and determined as being at increased risk of stroke or CVD complications, e.g., as compared to a control. The methods can include any therapeutic agents and / or procedures known in the art for the treatment and / or prevention of AF, silent AF, CVD or silent CVD, or ischemic disease such as stroke. An exemplary treatment includes anti-coagulant therapy. In some forms, the methods administer one or more therapeutic agents and / or procedures to the subject to reduce one or more symptoms of AF, silent AF, CVD or silent CVD or ischemic disease such as stroke in the subject. In other forms, the methods administer one or more therapeutic agents and / or procedures to the subject to prevent the onset or development of one or more symptoms of AF, silent AF, CVD or silent CVD or ischemic disease such as stroke in the subject. Therefore, in some forms, the methods administer one or more prophylactic agents and / or procedures to the subject. Effective reagents and amounts for the treatment and prevention of AF, CVD and / or stroke are known in the art. i. Anticoagulant therapy In some forms, the methods include one or more steps of administering anti-coagulant therapy to a subject identified as having or being at risk of having AF and / or CVD and / or thromboembolic stroke. For example, in some forms, the methods include treating a subject for AF, silent AF, CVD or silent CVD based on presence of one, or more than one, such as two, three, four, five, six, or more than six RIPLs in one or both eyes of the subject. The methods can include treating or preventing AF, silent AF, CVD or silent CVD using any therapeutic agents and / or procedures known in the art. 45667975.1 38 In some forms, the therapeutic agent is an anticoagulant, such as warfarin (COUMADIN®, JANTOVEN®) to help prevent the formation of blood clots within the small pulmonary arteries. Because anticoagulants prevent normal blood coagulation, they increase risk of bleeding complications. Exemplary treatment regimens include anti-coagulant therapy. Exemplary anti- coagulant therapeutic agents include apixaban (Eliquis), dabigatran (Pradaxa), edoxaban (Lixiana), rivaroxaban (Xarelto), warfarin (Coumadin), heparin and aspirin. Therefore, in some forms, the methods include identifying a subject as suitable for treatment with an anti- coagulant therapeutic agents include apixaban (Eliquis), dabigatran (Pradaxa), edoxaban (Lixiana), rivaroxaban (Xarelto), warfarin (Coumadin), heparin and aspirin. In some forms, the methods exclude the subject from treatment with an anti-coagulant therapeutic agents include apixaban (Eliquis), dabigatran (Pradaxa), edoxaban (Lixiana), rivaroxaban (Xarelto), warfarin (Coumadin), heparin and aspirin. Undesired side effects of anticoagulant therapy typically include an increased susceptibility to hemorrhages, such as passing blood in urine and / or feces, severe bruising, prolonged nosebleeds (lasting longer than 10 minutes), vomiting blood or coughing blood); sudden severe back pain; and difficulty breathing or chest pain; gastrointestinal diseases and disorders (including ulcers in the stomach or intestines); and Hemorrhagic stroke. ii. Other therapeutic options In some forms, the methods include one or more steps of administering one or more therapeutic agents to a subject identified as having or being at risk of having AF and / or CVD and / or thromboembolic stroke, for example, as part of a therapeutic or prophylactic treatment regime. In some forms, the methods identify a subject as likely to benefit from one or more specific treatment regimen, for example, anticoagulant therapy. In other forms, the methods identify a subject as likely to benefit from one or more specific treatments that is not anticoagulant therapy, for example, the administration of a statin. A non-limiting list of exemplary therapeutic agents includes anti-platelet therapy, vasoactive agents, such as catecholamines (e.g., adrenaline, noradrenaline, dopamine, dobutamine, isoprenaline, dopexamine), phosphodiesterase inhibitors (e.g., sildenafil, milrinone, amrinone, levosimendan), calcium sensitizers (e.g., levosimendan, glucagon), vasopressors (phenylephrine, metaraminol, ephedrine, vasopressin, steroids), digoxin Ca2+, thyroid hormone, epinephrine and methylene blue. 45667975.1 39 In some forms, the therapeutic agent is a blood vessel dilator (vasodilator) such as epoprostenol (FLOLAN®, VELETRI®). The drawback to epoprostenol is that its effects last only a few minutes. Epoprostenol is continuously injected through an intravenous (IV) catheter, for example via a small pump. Potential side effects of epoprostenol include jaw pain, nausea, diarrhea and leg cramps, as well as pain and infection at the IV site. In other forms, the additional therapeutic agent is iloprost (VENTAVIS®). Iloprost is inhaled six to nine times a day through a nebulizer. Because it's inhaled, it goes directly to the lungs. Side effects associated with iloprost include chest pain, often accompanied by a headache and nausea, and breathlessness. In other forms, the additional therapeutic agent is Treprostinil (TYVASO®, REMODULIN®, ORENITRAM®). Treprostinil is administered by injection, inhaled or taken orally up to four times a day. It can cause side effects such as a headache, nausea and diarrhea. In other forms, the additional therapeutic agent is Oral Uptravi (SELEXIPAG®). In some forms, the additional therapeutic agent is a Endothelin receptor antagonist, such as bosentan (TRACLEER®), macitentan (OPSUMIT®), and ambrisentan (LETAIRIS®). In some forms, the additional therapeutic agent is sildenafil (REVATIO®, VIAGRA®) or tadalafil (CIALIS®, ADCIRCA®). These drugs work by opening the blood vessels in the lungs to allow blood to flow through more easily. Side effects can include an upset stomach, headache and vision problems. In some forms, the additional therapeutic agent is a high-dose calcium channel blocker, such as amlodipine (NORVASC®), diltiazem (CARDIZEM®, TIAZAC®) and nifedipine (PROCARDIA®). Although calcium channel blockers can be effective, only a small number of people with pulmonary hypertension respond to them. In some forms, the additional therapeutic agent is a soluble guanylate cyclase (SGC) stimulator. Soluble guanylate cyclase (SGC) stimulators (ADEMPAS®) interact with nitric oxide and help relax the pulmonary arteries and lower the pressure within the arteries. These medications can sometimes cause dizziness or nausea. In some forms, the additional therapeutic agent is digoxin (LANOXIN®). Digoxin can help control heart rate and preclude arrhythmias. In some forms, the additional therapeutic agent is a diuretic (commonly known as water pills), to help eliminate excess fluid from the body and reduce the amount of work the heart has to do. They may also be used to limit fluid buildup in the lungs. In some forms, the additional therapeutic agent is oxygen. 45667975.1 40 In some forms, the methods identify a subject as likely to benefit from treatment with a combination of two or more different agents or procedures. The term “combination” or “combined” is used to refer to either concomitant, simultaneous, or sequential administration of two or more agents or procedures. Therefore, the combinations can be administered either concomitantly (e.g., as an admixture), separately but simultaneously (e.g., via separate intravenous lines into the same subject), or sequentially (e.g., one of the compounds or agents is given first followed by the second). For example, one or more anticoagulants can be administered on the same day, or a different day than the second active agent. In some forms, the second active agent can be administered on the first, second, third, or fourth day, following or before one or more anticoagulants. iii. Surgery In some forms, the methods identify a subject as likely to benefit from surgery, for example, heart and / or cardiovascular surgery. For example, in some forms, the methods identify a subject as likely to benefit from surgery to reduce the likelihood of a stroke. The current interventional approach to reduce stroke risk is the implantation of a preformed, rigid occlusion device that plugs the left atrial appendage space to prevent the development and embolization of clots. The techniques for mechanical occlusion are open heart surgery or percutaneous treatment, i.e., placement of a rigid structure to seal the opening, such as the WATCHMAN®, AMPLATZER®, and PLAATO® devices. In some forms, the additional therapeutic agent may aim to reduce CVD burden from atherosclerosis which may include vascular stent to the carotid artery, coronary artery or coronary artery bypass graft or stent for peripheral artery disease or aneurysm. In some forms, the additional interventions may include placement of a defibrillator, pacemaker or intravascular filter to reduce odds of thromboembolism. III. Diagnostic Systems Diagnostic systems that include the described methods are provided. In some forms, the methods determine a disease or disorder state of a subject based on the evaluation of the amount of RIPLs within one or both eyes of the subject. In some forms, the methods include repeating or reproducing one or more of the steps of the method, for example to provide multiple results. Therefore, in some forms, the methods include compiling or comparing two or more results and comparing results from other scans or systems that determine the presence and / or number of RIPLs in the eye of a 45667975.1 41 subject and optionally include determined a CHA2DS2-VASc Score, or a modified version, for the subject that includes RIPL data. In some forms, the methods are applied on images from commercially available OCT devices and / or are implemented into existing software or as a standalone software. Therefore, also disclosed are data structures used in, generated by, or generated from, the disclosed method. Data structures generally are any form of data, information, and / or objects collected, organized, stored, and / or embodied in a composition or medium. In an exemplary form, the detection and quantitation of RIPLs in one or both eyes of a subject and its association with a specific disease state and / or risk factor for a given subject, and / or additional information are stored in electronic form, such as in RAM or on a storage disk, as a type of data structure. The disclosed method, or any part thereof or preparation therefore, can be controlled, managed, or otherwise assisted by computer control. Such computer control can be accomplished by a computer controlled process or method, can use and / or generate data structures, and can use a computer program. Such computer control, computer controlled processes, data structures, and computer programs are contemplated and should be understood to be disclosed herein. In some forms, the methods include compiling and / or comparing a first set of results obtained from a subject with a second or further set obtained from the same subject, for example, obtained using the same or different scan data, obtained at the same or different times. In some forms, the methods include compiling and / or comparing the results obtained from a first scan of the eye with those obtained from a second of further scan of the eye obtained at different times, such as with a time course of one day, one week, one month, or one year or more than one year. The different scans can be from the same or different equipment, and / or obtained at the same or different location, at the same or different times relative to one another. In some forms, the methods include compiling and / or comparing the results obtained from a OCT scan of the eye(s) with those obtained from the same subject, for example at the same or different times. In some forms, the methods determine a disease or disorder state of a subject based on the evaluation of the amount of RIPLs from the subject over a time course of between one day and ten or more years, such as a course of one week, one month, three months, six months, one year, two years, five years or ten years. In some forms, the methods determine a disease or disorder state of a subject based on the evaluation of the amount of RIPLs within the eye of a healthy subject and / or a subject with CVD. 45667975.1 42 The disclosed methods can further involve the use of a computer system to compare levels of the one or more of the RIPLs in the eye(s) of a subject to control values. For example, the computer system can use an algorithm to compare levels of two or more factors (such as CHA2DS2-VASc Score (or any CVD risk factors) and RIPLs for a given subject, and provide a score representing the risk of disease onset based on detected differences. This algorithm can in some forms weigh multiple parameters. For example, in some forms, the algorithm gives weight depending on which biomarker demonstrates differences, e.g., more weight to differences in RIPLs over other factors or biomarkers. In some forms, the algorithm weighs the extent of elevation or decrease in biomarkers levels compared to the control. For example, a 50% reduction in biomarker levels may be weighted more than a 20% reduction in the same biomarker. In other forms, the algorithm gives weight to differences in biomarker levels for a combination of biomarkers. In other forms, the algorithm gives weight to whether RIPLs are present in one or both eyes. Therefore, also provided is an apparatus for use in detecting CVD, or any other disease or disorder associated with the increased or abnormal presence of RIPLs in the eyes of a subject that includes an input means for entering biomarker level values from a sample of the subject, a processor means for comparing the values to control values, an algorithm for giving weight to specified parameters, and an output means for giving a score representing the risk of disease onset. The described systems and methods will be better understood in view of the following numbered paragraphs. 1. A method including: (i) determining the presence or absence of retinal ischemic perivascular lesions (RIPLs) in an eye of a subject; and (ii) selecting the subject for treatment for increased risk of stroke when the determining in step (i) identifies one or more RIPLs in an eye of the subject. 2. The method of paragraph 1, wherein the subject has, or has been identified as being at increased risk of having, atrial fibrillation (AF) and / or other cardiovascular disease (CVD). 3. The method of paragraph 1, wherein the subject does not have atrial fibrillation (AF) and / or other cardiovascular disease (CVD). 4. The method of any one of paragraphs 1-3 further including, prior to step (ii), determining a risk factor (CHA₂DS₂-VASc) score for the subject, 45667975.1 43 wherein the risk factors include one or more of Congestive heart failure, Hypertension, Age≥75 years, Diabetes mellitus, Prior Stroke or TIA or Thromboembolism, Vascular disease, Age 65–74 years, and Sex category. 5. The method of paragraph 4, further including: (iii) administering anti-ischemia treatment and / or anti-stroke treatment to the subject selected in step (ii) if the CHA₂DS₂-VASc score is 1 or above. 6. The method of paragraph 1 further including: (iii) administering anti-ischemia treatment and / or anti-stroke treatment to the subject selected in step (ii). 7. The method of paragraph 5 or 6, wherein the treatment includes anti-coagulation therapy. 8. The method of paragraph 7, wherein the anti-coagulation therapy includes administering an anticoagulant agent to the subject, wherein the anticoagulant agent is selected from the group including warfarin, apixaban, dabigatran, edoxaban, rivaroxaban, heparin, and aspirin. 9. The method of any one of paragraphs 5 to 8, wherein the treatment includes surgery. 10. The method of paragraph 1 further including excluding the subject from a treatment or refraining from treating the subject when the subject is not selected in step (ii). 11. The method of paragraph 4 further including excluding the subject from a treatment or refraining from treating the subject when the CHA₂DS₂-VASc score is below 1. 12. The method of paragraph 1 further including (a) discontinuing anti-ischemia treatment and / or anti-stroke treatment or (b) refraining from further anti-ischemia treatment and / or anti-stroke treatment of the subject that is or had been receiving anti-ischemia treatment and / or anti-stroke treatment. 13. The method of any one of paragraphs 1 to 10, wherein the determining in step (i) includes imaging one or both eyes of the subject. 14. The method of paragraph 13, wherein the imaging includes optical coherence tomography (OCT). 15. The method of paragraph 14, wherein the OCT is spectral domain optical coherence tomography (SD-OCT). 16. The method of paragraph 14, wherein the OCT is time domain optical coherence tomography (TD-OCT). 45667975.1 44 17. The method of paragraph 14, wherein the OCT is swept source optical coherence tomography (SS-OCT). 18. The method of paragraph 14, wherein the OCT is visible light optical coherence tomography (VL-OCT). 19. The method of paragraph 12, wherein the OCT is macular optical coherence tomography (OCT). 20. The method of any one of paragraphs 1 to 19, wherein the determining in step (i) includes: (a) providing an OCT scan image of an eye of a subject; (b) identifying a cross section in the image corresponding to the retina which is flanked by the vitreous cavity and choroid; (c) fitting a polynomial curve to top points, bottom points, and A center curve passing through the midpoint of the cross section to provide a fitted image; (d) optionally processing the fitted image to remove noise to provide a processed image, wherein the processing optionally includes one or more of thresholding, image dilation, and erosion; (e) filtering the fitted and / or processed image to accentuate horizontal lines using specified kernels to provide a filtered image, optionally, wherein the filtering includes Median filtering and / or Kernel filtering; (f) re-tracing lines and the center curve in the filtered image to provide a retraced image; and (g) identifying and / or quantifying RIPLs on the center lines and curve, optionally wherein the identifying includes normalization of slopes compared to reference line. 21. The method of paragraph 20 further including: (h) (1) annotating the identified RIPLs on the original image, and / or (2) quantitating and / or annotating the total number of RIPLs per scan; and / or (i) quantitating and / or recording a total number of RIPLs per eye and / or per subject. 22. The method of any one of paragraphs 1 to 21, wherein the determining in step (i) includes detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) via visual inspection of images of the inner retinal layer (IRL) of an eye of the subject. 23. The method of paragraph 22, wherein the visual inspection includes duplicate, triplicate, or further inspections. 45667975.1 45 24. The method of paragraph 23, wherein the duplicate, triplicate, or further inspections are carried out by two or more different inspectors. 25. The method of any one of paragraphs 1 to 21, wherein the determining in step (i) includes detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) in the eye of the subject via a computer-based system for the automated analysis of images of the inner nuclear layer, outer nuclear layer, inner plexiform layer, and outer plexiform layer of an eye of the subject. 26. The method of paragraph 25, wherein the computer-based system for the automated analysis of images includes: (a) providing input data to a computer processor operably connected to means for computing and displaying output data, wherein the input data includes one or more images of the retinal layer (inner nuclear layer, outer nuclear layer, inner plexiform layer and outer plexiform layer) of an eye of the subject, or data derived therefrom, wherein the output data includes the number of RIPLs detected within the one or more images; and (b) computing the output data. 27. The method of paragraph 26, wherein the input data includes a computer-readable media. 28. The method of any one of paragraphs 1 to 19, wherein the detecting in step (i) includes determining an RIPL index for a scan, wherein the RIPL index quantifies total positive deviation of interface between an outer plexiform layer (OPL) and an outer nuclear layer (ONL) above an average position for each B-scan. 29. The method of paragraph 28, wherein determining the total positive deviation of the OPL / ONL interface above average includes: (a) providing one or more OCT B-scan images of an eye of a subject, wherein a scan image includes a multiplicity of pixel intensities, wherein a pixel intensity of the multiplicity of pixel intensities includes a value, and wherein the multiplicity of pixel intensity values are disposed as columns and rows throughout the image; (b) processing the B-scans to reduce noise and highlight internal limiting membrane (ILM) and retinal pigment epithelium (RPE) layers in the images to provide a processed scan having columns of pixel intensity, 45667975.1 46 optionally wherein the highlighting includes one or more processing steps of erosion, dilation, and / or thresholding; (c) segmenting the ILM and RPE, wherein the segmenting includes iterating down each column of pixel intensity values and up each column of pixel intensity values, respectively, until there is a change in pixel intensity value of at least a first threshold amount; (d) excluding scans of fovea, wherein the excluding includes determining an average central retinal thickness and removing B-scans having a central retinal thickness that is a second threshold amount less than the average retinal thickness; (e) locating the darkest pixel in each column to identify the outer plexiform layer (OPL), wherein the darkest pixel has the greatest intensity value; (f) segmenting out the outer nuclear layer (ONL) / OPL interface, wherein the segmenting includes iterating up each pixel in each column until a third threshold increase in pixel intensity; (g) calculating the pixel height between the RPE and ONL / OPL interface for each column; (h) optionally excluding heights several standard deviations above the mean to reduce errors in ONL / OPL interface detection; (i) obtaining positive deviations of the ONL / OPL interface, wherein the obtaining includes determining the average height of the ONL / OPL interface above the RPE, and subtracting the average height of the ONL / OPL interface above the RPE from the height at every column above the average; and (j) summing the positive deviations to provide the total volume of deviation of the OPL / ONL interface above average for each B-scan. 30. The method of paragraph 29, wherein step (b) further includes removing shadowing from the images, wherein the removing includes providing a best fit spline having a single pixel thickness to trace the RPE; tracing the OPL, wherein tracing the OPL includes increasing the thickness of the best fit spline from 1 pixel in thickness to up to about 7 pixels in thickness and moving the spline up by a proportion of the average distance between the RPE and ONL / OPL interface; calculating the average pixel intensity for each column in the 7-pixel height RPE slice and 7-pixel height OPL slice; and 45667975.1 47 excluding a column from the OPL slice to account for shadowing when the RPE slice average pixel intensity are a fourth threshold value below average. 31. The method of paragraph 30, further including: identifying and counting multiple consecutive columns in the OPL slice having pixel intensity above the fourth threshold value as a single RIPL; and excluding multiple consecutive dark columns above a fifth threshold value as being attributed to fovea. 32. The method of any one of paragraphs 1 to 19, wherein RIPLs are identified from an OCT scan utilizing a computer program run on a computer system, wherein the program is configured to identify RIPLs in the OCT scan by segmentation of the outer plexiform layer (OPL) which follows the curvature of the retinal pigment epithelium. 33. The method of any one of paragraphs 28 to 32, wherein the method is repeated one or more times to sequentially identify RIPLs in scans from two or more subjects or pools of subjects. 34. The method of paragraph 32, wherein the repeated method is carried out sequentially by a computer as a first and second or further iterations, wherein the computer implements a machine-learning system to increase the efficacy of the method in a second or further iteration as compared with a first or further iteration. 35. The method of any one of paragraphs 1 to 34, wherein the method further includes, prior to step (ii), one or more steps to determine whether the subject has AF. 36. The method of any one of paragraphs 1 to 35, wherein the AF is asymptomatic AF or silent AF. 37. The method of any one of paragraphs 1 to 36, wherein the subject has a personal or family history of diseases or disorder selected from the group including AF, cardiovascular disease, hypertension, carotid disease, diabetes, hypercholesterolemia, hyperlipidemia, stroke, and ischemic disease. 38. The method of any one of paragraphs 1 to 37, wherein the subject has an implantable cardiac device, and / or has had recent cryptogenic stroke or transient ischemic attack (TIA), or vascular dementia. 39. The method of any one of paragraphs 1 to 38, wherein the subject has a personal or family history of diseases or disorder selected from the group including retinopathy, diabetic retinopathy, hypertensive retinopathy, ischemic retinopathy, or macular edema. 45667975.1 48 40. The method of any one of paragraphs 1 to 19, wherein the determining in step (i) comprises: (a) providing an OCT scan image of an eye of a subject; (b) identifying a cross section in the image corresponding to the retina which is flanked by the vitreous cavity and choroid; and (c) identifying the presence of RIPLs within the cross section in the image, optionally wherein the cross section comprises the inner nuclear layer, the outer plexiform layer and the outer nuclear layer of the eye. 41. The method of paragraph 40, wherein identifying the presence of RIPLs within the cross section in step (c) comprises: (I) fitting a polynomial curve to top points, bottom points, and A center curve passing through the midpoint of the cross section to provide a fitted image; (II) optionally processing the fitted image to remove noise to provide a processed image, wherein the processing optionally comprises one or more of thresholding, image dilation, and erosion; (III) filtering the fitted and / or processed image to accentuate horizontal lines using specified kernels to provide a filtered image, optionally, wherein the filtering comprises Median filtering and / or Kernel filtering; (IV) re-tracing lines and the center curve in the filtered image to provide a retraced image; and (V) identifying and / or quantifying RIPLs on the center lines and curve, optionally wherein the identifying comprises normalization of slopes compared to reference line. 42. The method of paragraph 40, wherein identifying the presence of RIPLs within the cross section in step (c) comprises: visual inspection of the cross section images. 43. The method of paragraph 42, wherein the presence of RIPLs is indicated by an observed expansion in the outer nuclear layer hypo-reflectivity. 44. The method of paragraph 42 or 43, wherein the visual inspection comprises duplicate, triplicate, or further inspections. 45. The method of paragraph 44, wherein the duplicate, triplicate, or further inspections are carried out by two or more different inspectors. 45667975.1 49 46. The method of any one of paragraphs 40-45, further comprising annotating the identified RIPLs on the original image, and / or quantitating and / or annotating the total number of RIPLs per scan; and / or quantitating and / or recording a total number of RIPLs per eye and / or per subject. 47. The method of any one of paragraphs 1 to 19, wherein the determining in step (i) comprises detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) via visual inspection of one or more image(s) of the inner retinal layer (IRL) of an eye of the subject, optionally wherein (a) the presence of RIPLs is indicated by an expansion in the outer nuclear layer hypo-reflectivity; and / or (b), wherein the visual inspection comprises duplicate, triplicate, or further inspections, optionally wherein the duplicate, triplicate, or further inspections are carried out by two or more different inspectors. The described systems and methods will be better understood in view of the following Examples. EXAMPLES Example 1: Retinal Ischemic Perivascular Lesions (RIPLs) are a Biomarker of Cardiovascular Disease Aim To analyze results of cardiovascular workup from a consecutive cohort of patients with no preexisting cardiovascular disease in whom Retinal Ischemic Perivascular Lesions (RIPLs) were noted during the retinal examination. Methods The study adhered to the tenets of the Declaration of Helsinki. Institutional Review Board approval was obtained with a waiver of informed consent for retrospective chart review. Consecutive patients from the retinal clinic were identified in whom: (i) RIPLs were detected on SD-OCT taken for various clinical indications; and (ii) who did not have any evidence of underlying retinopathy. 45667975.1 50 A RIPL was defined by the SD-OCT presence of focal atrophy of the inner nuclear layer (INL) associated with the secondary expansion of the outer nuclear / Henle’s fiber layer, leading to a wavy appearance of the middle retinal layers as described by Long et al., EclinMed.2021;33:100775. Inclusion criteria included a high quality 6 x 6 mm SD-OCT volume macular scan consisting of 49 B-scans (cross-sections). The SD-OCT scan is part of the routine workflow at our institution for patients referred for retinal examination. Exclusion criteria were: (i) lack of follow up at our institution, (ii) presence of macular edema, or any evidence of central retinal artery or vein occlusion, (iii) presence of posterior uveitis, (iv) presence of pars plana vitrectomy, (v) evidence of intravitreal injection, or (vi) evidence of pan-retinal photocoagulation. At the time of examination, individuals with RIPLs but without a prior history of cardiovascular disease other than essential HTN were advised to follow up with their primary care physician or cardiologist for an age-appropriate cardiovascular workup at their physician’s discretion. Those who did not seek a medical follow-up were excluded. Results 36 patients were identified with incidental RIPL identification on SD-OCT. An exemplary RIPL detected on spectral-domain (SD) OCT scans is shown in Figs.1A-1G, showing an infrared image of the retina of an eye with no appreciable abnormalities, and also an eye with an RIPL. Indications for retinal examination included floaters, posterior vitreous detachment, visual disturbance, screening for diabetic retinopathy, and referral for retinal examination following cataract surgery. A total of 25 patients were excluded; ten patients were excluded due to the presence of preexisting cardiovascular disease, including coronary artery disease, congestive heart failure, and atrial fibrillation, as well as prior adverse cardiovascular events like pulmonary embolism, deep venous thrombosis, and stroke; and a total of 15 individuals chose not to pursue further cardiovascular workup or pursued it outside our institution. 45667975.1 51 Of the 11 patients that were included, 5 were men and 6 were women. Ages ranged from 44 to 80 years. The 11 patients with no prior history of cardiovascular disease other than essential HTN followed up with their primary care physician or cardiologist for age appropriate cardiovascular workup. Of these, newly diagnosed cardiovascular disease was identified in 8 individuals (72.7%), including multivessel coronary artery disease, significant carotid artery stenosis, soft carotid plaque, reduced cardiac ejection fraction, patent foramen ovale, cerebral infarction, subclavian steal syndrome, and undiagnosed or poorly controlled hypertension (HTN). Two patients subsequently underwent invasive procedures; in case 1, the patient required coronary artery bypass grafting after being found to have three-vessel disease on coronary angiography, which was prompted by an abnormal echocardiogram; in case 2, the patient underwent carotid artery stent placement after observation of multifocal atherosclerotic disease on head and neck computed tomography angiography. Three patients initiated new medical therapy for their conditions, including antihypertensives and anticoagulants. Two patients were advised to undergo periodic monitoring. Summary These findings highlight the efficacy of detecting RIPLs to uncover underlying cardiovascular disease in a clinical setting. Previously a retrospective series of 160 individuals demonstrated that the presence of 1, 2, or 3 RIPLs was associated with an odds ratio of 2.34, 4.17, and 5.34 of having cardiovascular disease, respectively, after adjusting for age, sex, and smoking status. Therefore, those with 3 RIPLs are more than 5 times more likely to have CVD than those with no RIPLs. Retinal ischemic perivascular lesions, which are anatomical biomarkers of retinal ischemia, have also been documented in retrospective studies in individuals with HTN and in the fellow eyes of patients with unilateral retinal vein occlusion. As a legacy of paracentral acute middle maculopathy, these lesions represent INL infarcts and may represent the earliest evidence of macular ischemia. In this series, we present data that further supports the role of RIPL identification in detecting coexisting cardiovascular disease with the goal of preventing major systemic cardiovascular events. Besides the obvious clinical benefit in reducing morbidity and mortality, early intervention could also reduce the financial burden associated with managing complications such as stroke and heart attack. Retinal ischemic perivascular lesion detection with SD-OCT is a simple, fast technique that is inexpensive, noninvasive, 45667975.1 52 already part of the routine ocular examination, and widely used in ophthalmology and optometry clinics. In conclusion, the findings strongly suggest that RIPLs provide an additional biomarker to predict the existence of subclinical cardiovascular disease, and that the number of RIPLs increases with the likelihood of CVD in subjects having subclinical indications. Example 2: Retinal Ischemic Perivascular Lesions (RIPLs) in Individuals with Atrial Fibrillation Aim In the present study, whether RIPLs are associated with atrial fibrillation (AF), independent of underlying ischemic heart disease. Identifying whether this association exists is important, as it may help inform practitioners on the appropriate medical evaluation needed for a patient who presents to ophthalmology clinic and is incidentally found to have RIPLs on OCT. Methods This was a case-control study that adhered to the tenets of the Declaration of Helsinki and was conducted in accordance with the regulations of the Health Insurance Portability and Accountability Act. Institutional Review Board (IRB) approval for retrospective chart review was obtained. The need for informed consent was waived because of the retrospective nature of the study. Individuals who were between 50 and 90 years of age and who underwent macular SD-OCT scanning between January 1, 2018 and December 31, 2020 for various clinical indications were identified. The majority of patients evaluated in ophthalmology clinics have an OCT scan performed as part of the routine clinical evaluation and workflow. Exclusion criteria included: (i) any evidence of retinal vascular occlusion (including branch and central retinal artery or vein occlusion); (ii) any history of retinal laser; (iii) any history of intravitreal injection; and (iv) any history of pars plana vitrectomy. Subjects with atrial fibrillation were identified based on the ICD-10 diagnostic code I48, who also had a cardiac rhythm monitor (CPT code 93224-9 or 33285-6) recorded. These subjects were selected as cases. Frequency-matching was used to select controls. Controls included individuals who were age and sex-matched from the same pool, but without a diagnosis of atrial fibrillation. 45667975.1 53 The medical records of both the atrial fibrillation and control groups were individually reviewed to confirm the presence or absence of atrial fibrillation. Additionally, using electronic health record review, baseline characteristics were recorded including age, sex, smoking status, diagnosis of hypertension, diabetes mellitus, coronary artery disease, carotid stenosis, stroke, and myocardial infarction. The macular SD-OCT scans were reviewed, and subjects with any macular pathology, other than RIPLs, were excluded. Poor quality scans, in which the individual layers of the retina were indiscernible due to media opacity, were also excluded. RIPLs grading All SD-OCT scans that were included and reviewed in this study were acquired on a Spectralis SD-OCT machine (Heidelberg Engineering, Heidelberg, Germany) as part of the routine retinal examination. RIPLs were identified using the previously published criteria. Briefly, individual b-scans were reviewed by three independent and masked observers, and a RIPL was defined by the presence of focal atrophy or thinning of the inner nuclear layer with a compensatory expansion of the outer nuclear layer. A consensus between all three observers was required to identify a RIPL. Subjects were classified into two groups, those with and without RIPLs. Statistical Analysis Baseline characteristics of subjects are described using mean and standard deviation (SD) for continuous variables and frequency with percent for categorical variables. Demographics were compared for cases and controls using Pearson’s Chi-Squared Test for categorical variables and Student’s t-test for continuous variables. Univariable and multivariable logistic regression models were utilized to evaluate the relationship between RIPLs and atrial fibrillation. Covariates included: age, sex, smoking status, hypertension, diabetes mellitus, coronary artery disease, carotid stenosis, stroke, and myocardial infarction. All statistical analyses were performed using SPSS Version 28.0 (Armonk, NY). Results A total of 106 individuals with AF who met all inclusion and exclusion criteria were identified and 91 subjects without atrial fibrillation as controls. Baseline characteristics for cases and controls are shown in Table 1. The mean age was 72.5 (SD 7.8) years and 70.6 (SD 7.4) years for cases and controls, respectively. Females included 47.2% and 48.4% of the case and control groups, respectively. There was no difference between both groups in terms of age and sex distribution. A 45667975.1 54 significantly greater proportion of cases had hypertension (77.4%) and stroke (12.3%) as compared to controls (66.2% and 1.1%, respectively). Additionally, 37.7% of cases had coronary artery disease, as compared to 19.8% of controls (Table 3). There was no difference in prevalence of diabetes mellitus between the two groups. The percentage of subjects with RIPLs was higher in the atrial fibrillation group compared to the control group (57.5% vs 37.4%, p = 0.005). Table 3. Baseline characteristics of the study cohort. Mean and standard deviation reported for age. Frequency and percentage reported for other variables. Abbreviations: RIPLs = retinal ischemic perivascular lesions. In a multivariable logistic regression model adjusted for age and sex, the presence of RIPLs was significantly associated with atrial fibrillation with an odds ratio (OR) of 2.19 (95% CI 1.21 – 3.97) (p=0.009). After adjusting for additional covariates including smoking history, hypertension, diabetes mellitus, coronary artery disease, carotid stenosis, stroke, and myocardial infarction, the presence of RIPLs continued to be associated with atrial fibrillation with an OR of 1.91 (95% CI 1.01 – 3.59) (Table 4). 45667975.1 55 Table 4. Multivariable logistic regression model evaluating the relationship between presence of one or more RIPLs and a diagnosis of atrial fibrillation. Abbreviations: RIPL(s) = retinal ischemic perivascular lesion(s); OR = odds ratio; CI = confidence interval. Summary In this case-control study, a significant association between RIPLs and atrial fibrillation that was independent of underlying ischemic heart disease was identified. Atrial fibrillation is a cardiac arrhythmia that is prevalent in the adult population and increases in prevalence with age. The lifetime risk of atrial fibrillation is estimated to be about 1 in 4 individuals. Because of the rising proportion of elderly persons within the population, it is projected that the prevalence of atrial fibrillation is likely to increase 2.5- fold over the next 50 years. A study from 2011 estimated the national incremental cost of atrial fibrillation to be $26 billion, accounting for 3.2 million hospital days. Individuals with atrial fibrillation are at a higher risk for developing thromboembolic events including cerebral strokes, as well as other co-morbidities such as heart failure, cognitive decline, and decline in kidney function. Approximately 0.7 million (13%) of the 5.3 million cases of atrial fibrillation in the United States may be undiagnosed. 45667975.1 56 In this study, we identified that a subclinical biomarker of retinal ischemia, RIPL, is significantly associated with atrial fibrillation, independent of other comorbidities. The pathogenesis of RIPL due to atrial fibrillation may be the result of hypoperfusion of the deep retinal capillary plexus, which is at greater risk of ischemia compared to the superficial capillary plexus. Retinal capillary hypoperfusion may develop secondary to microemboli formation or reduced cardiac output. Previous work established a relationship between RIPLs and ischemic heart disease. This study establishes a relationship between RIPLs and atrial fibrillation, a major source of emboli formation. It is not suggested that the identification of this biomarker would be more feasible than screening for atrial fibrillation using electrocardiogram; however, if RIPLs are identified in an individual without known cardiovascular disease on routine examination with an ophthalmologist, these data would suggest that a broad cardiac evaluation may be warranted. This is further supported by previous findings which also show an independent association between RIPLs and ischemic heart disease. Conversely, a patient with atrial fibrillation who is found to have RIPLs, may not necessarily benefit from additional cardiovascular workup. In conclusion, it has been established that RIPLs, which are anatomical biomarkers of subclinical ischemia, are significantly associated with atrial fibrillation, independent of other underlying cardiovascular disease. This information can guide the medical management of patients with RIPLs incidentally detected on macular OCT. Example 3: Retinal Ischemic Perivascular Lesions (RIPLs) are Associated with Stroke in Individuals with Atrial Fibrillation Aim Atrial fibrillation (AF) is a prevalent cardiac arrhythmia that substantially increases the risk of stroke. To determine whether retinal ischemic perivascular lesions (RIPLs) are associated with subclinical retinal ischemia, and stroke in individuals with AF, after adjusting for the conventional risk stratification tools used in clinical practice, the presence of RIPLs in the eyes of AF and stroke patients was assessed. Whether this relationship is independent of underlying cardiovascular risk factors was also assessed within a cross-sectional cohort of individuals with AF. Whether the addition of RIPLs to the CHA2DS2-VASc score would provide improved prognostic utility in predicting strokes was also examined. 45667975.1 57 Methods This is a retrospective, cross-sectional study which adhered to the Tenets of Helsinki. Institutional Review Board (IRB) approval was obtained. subjects were identified through Epic’s Slicer Dicer tool using ICD10 codes of AF (I48.0 - I48.4, I48.91 - I48.92) and presence of a macular OCT. Subjects aged 50 to 90 years diagnosed with AF who had undergone optical coherence tomography (OCT) with a documented electrocardiogram or heart monitor, were selected. RIPLs were identified on macular OCT scans by three independent graders. Electronic medical charts were individually reviewed to confirm the diagnosis of atrial fibrillation and collect pertinent clinical information including age, sex, anticoagulation use, presence of hypertension, diabetes mellitus, vascular disease, congestive heart failure, and stroke or transient ischemic attack (TIA). Vascular disease was defined as prior myocardial infarction, peripheral artery disease or aortic plaque. Exclusion criteria included: (i) the presence of retinal artery or vein occlusions; (ii) diabetic retinopathy; (iii) history of pars plana vitrectomy; and (iv) history of retinal laser photocoagulation or intravitreal injections. Additional exclusion criteria upon review of OCT scans were: (v) the presence of macular pathology such as significant epiretinal membrane; (vi) advanced macular degeneration, and / or choroidal neovascularization; and (vii) poor quality scans in which the discrete layers of the retina were not discernible. OCT scans were reviewed for presence of RIPLs by three independent and masked observers. RIPLs were identified upon review of b-scans as evident by focal thinning in the inner nuclear layer (INL) that is coupled by expansion in the outer nuclear layer (Henle’s fiber layer) hypo-reflectivity. A consensus agreement between all three observers was required in order to document presence of RIPLs within each subject. Statistical analysis This study reports the baseline clinical variables of those individuals with and without RIPLs. Categorical variables are reported as frequency and percent. Continuous variables are reported as mean and standard deviation. Differences between the groups were determined 45667975.1 58 using Student’s t-test for continuous variables and Fisher’s Exact Test for categorical. Multivariable logistic regression models were utilized to evaluate the relationship between presence of RIPLs and stroke within individuals with AF. Variables deemed important to include in these models a priori were: age, sex, a diagnosis of diabetes mellitus, hypertension vascular disease, and congestive heart failure. Next, because this is a cross-sectional study evaluating stroke as the outcome, a modified CHA2DS2-VASc score was calculated (“CHA2D-VASc”), in which history of stroke, transient ischemic attack or thromboembolism were excluded as a risk factor. It was determined to do this, since the cross-sectional nature of the study prohibited determining the timing of strokes. A multivariable logistic regression was then utilized to evaluate the relationship of RIPLs with stroke, after adjustment for modified CHA2D-VASc score. The area under the curve (AUC) was calculated for two models: modified CHA2D-VASc score alone and modified CHA2D-VASc score + RIPLs. Models were compared using the Likelihood Ratio Test. All statistical analyses were performed using R Version 4.2.2 *** (Vienna, Austria). Results 178 individuals with AF who met all inclusion and exclusion criteria were identified. Baseline demographics and clinical variables are presented in Table 5. Within the cohort, 73 subjects (41%) had RIPLs detected on OCT scan. There were 30 subjects overall (16.9%) who had a history of stroke or TIA. Subjects with stroke or TIA had significantly higher prevalence of hypertension (90% vs.71.6%, p=0.038), congestive heart failure (43.3% vs. 24.3%, p=0.044), and RIPLs (63.3% vs.36.5%, p=0.008) compared with those without stroke or TIA. Overall, 122 subjects (68.5%) were on anticoagulation, and there was no significant difference in anticoagulation use in those who did (66.7%) or did not (68.9%) have stroke or TIA (Table 5). Table 5: Characteristic of Study Cohort No stroke or Stroke or TIA Overall p-value RIPLs consensus 54 (36.5) 19 (63.3) 73 (41.0) 0.008 Congestive Heart 36 (24.3) 13 (43.3) 49 (27.5) 0.044 attack. The numbers represent the mean and standard deviation for continuous variables (age) and frequency with percent for categorical variables. For example, females included 50% of the no stroke group (74 / 148) and 50% of the stroke group (15 / 30). Table 6: Baseline demographic characteristics of subjects. No stroke or TIA Stroke or TIA Overall p-value N 148 N 30 N 178 In a univariate analysis, individuals with RIPLs had an odds ratio (OR) of 3.01 (1.33, 6.79) for experiencing a stroke (p= 0.008). Using multivariable logistic regression models adjusted for age, sex, anticoagulation use, presence of hypertension, diabetes mellitus, vascular disease, and congestive heart failure, individuals with RIPLs had an OR 2.88 (95% CI 1.23, 6.98) for a diagnosis of stroke or Transient Ischemic Attack (TIA), respectively (p=0.016) (Table 6). 45667975.1 60 Table 7: Multivariable logistic regression model evaluating the relationship of presence of RIPLs with stroke in subjects with atrial fibrillation. Variable Estimate (95% CI) p-value Presence of RIPLs 2.88 (1.24, 6.98) 0.016 Finally, the association between RIPLs and stroke was tested in a model which also included modified CHA2D-VAScscore and found that both RIPLs and modified CHA2D- VASc score were significantly associated with stroke. In this model, each additional CHA2D- VASc point was associated with Odds Ratio (OR) 1.40 (1.03, 1.91) of stroke, and presence of RIPLs was associated with OR 2.87 (1.26, 6.55) of stroke (p=0.03, 0.01 respectively, Table 8). Table 8: Multivariable logistic regression model evaluating the relationship of presence of RIPLs with stroke, after adjusting for modified CHA2D-VASc score. Variable Estimate (95% CI) p-value failure, Hypertension, Age ≥ 75, Diabetes mellitus, Vascular disease, Age 65-74 years, Sex category. This model, which incorporated both CHA2D-VASc and RIPLs, was superior to the model utilizing modified CHA2D-VASc alone based on the Likelihood Ratio test (p=0.01). Accordingly, combining presence of RIPLs with modified CHA2D-VASc score in an ROC analysis, the area under the curve (AUC) in determining stroke was 0.70. This AUC was an increase from 0.62 when using modified CHA2D-VASc alone (Figure 3). 45667975.1 61 Table 9: ROC Curves for modified CHADS Vasc score alone and for modified CHADS Vasc + RIPLs diagnosis. Discussion In this retrospective, cross-sectional study, it was established that presence of RIPLs was significantly associated with stroke in individuals with atrial fibrillation. This relationship remained highly significant, even after adjustment for other established risk factors for stroke. The addition of OCT assessment of RIPLs to a CHA2D-VASc score significantly improved the AUC of the model for discriminating stroke. These findings suggest that the presence of subclinical retinal ischemia in patients with atrial fibrillation may indicate an increased risk of stroke and provide additional information to guide physicians in balancing the risks vs. benefits of initiation of anticoagulation therapy. RIPLs are reminiscent of prior microinfarcts arising from ischemia or microemboli. This study confirmed that RIPLs are associated with stroke in patients with AF. This association exists because of the shared pathophysiological mechanisms for the development of both RIPLs and stroke in patients with AF. The study demonstrates that patients with atrial fibrillation with RIPLs are at higher risk for stroke development as compared to those without RIPLs. AF is a common disease in older adults, and the number of patients living with AF is expected to increase 2.5-fold over the next 50 years, as the elderly population grows. The national incremental AF cost is estimated to be between $6 to $26 billion. Patients with AF 45667975.1 62 have a four times increased risk of mortality as compared to the general population, and the leading cause of death in individuals with AF is cerebral infarction. As such, it becomes very important to be able to optimize management of patients with AF and to have a tailored approach to determining the risk-benefit ratio of initiating anticoagulation therapy. For patients with AF and a CHA2DS2-VASc score of two or more who were not receiving anticoagulation therapy, the top five reasons physicians cited for no anticoagulation were: 1) low AF burden or successful rhythm control 2) patient refusal 3) perceived low risk of stroke 4) fall risk and 5) high bleeding risk. In this retrospective review, RIPLs, as detected with OCT, are identified as being associated with increased risk of stroke in patients with AF. RIPLs can be readily identified with non-invasive OCT scans that are likely being performed in patients with AF, who may be following with an ophthalmologist because of comorbidities such as diabetes or hypertension. The presence or absence of RIPLs may help guide the conversation on anticoagulation between patients with AF and their providers in situations where the estimated risks or benefits to be derived are uncertain. In conclusion, in a cross-sectional, retrospective analysis of individuals with AF followed at two academic centers, RIPLs, an OCT marker of subclinical retinal ischemia, were identified as being significantly associated with stroke. Of the 178 individuals with AF identified, RIPLs were observed in 41%, and 16.9% experienced stroke or TIA. There was no significant difference in anticoagulation use between those who did and did not have stroke or TIA. After adjusting for age, sex, anticoagulation use, presence of hypertension, diabetes mellitus, vascular disease, and congestive heart failure, presence of RIPLs was significantly associated with increased risk of stroke or TIA (OR 2.88 [1.23, 6.98], p=0.016). A model incorporating both a modified CHA2DS2-VASc score and presence of RIPLs on OCT significantly outperformed a model relying solely on the modified CHA2DS2-VASc score (p=0.01). The addition of RIPLs to a modified CHADS Vasc score increased the AUC for diagnosing stroke from 0.62 to 0.70. Macular OCT scans are a non-invasive, easily accessible test. Including the OCT assessment of RIPLs along with the CHA2D-VASc score significantly improved the discrimination of stroke. Consequently, the identification of RIPLs on scans in patients with atrial fibrillation may help guide decision making on anti- coagulation therapy, particularly in those patients who may be at higher risk of complications from treatment. Assessment of RIPLs with OCT improves discrimination of stroke in 45667975.1 63 individuals with AF beyond the conventional CHA2DS2-VASc score. Future prospective studies are needed to determine the potential utility of RIPLs in refining AF management approaches. Example 4: Automated Detection of RIPLs Computer vision heuristic model to calculate a RIPL index An exemplary automated method for detecting RIPLs in OCT scan images utilizes a computer vision heuristic model to calculate a RIPL index for each OCT volume (B-scans). The RIPL index served as a measure of the total positive deviation of the ONL / OPL interface above its average position. For each B-scan, erosion, dilation, and thresholding were used to reduce noise and highlight pertinent layers. The ILM and RPE were then segmented out by iterating down each column and up each column respectively until there was a threshold change in pixel intensity. B-scans with a central retinal thickness a certain amount less than the average retinal thickness were excluded as a method for excluding scans of the fovea. The darkest pixel was then located in each column to identify the OPL. The ONL / OPL interface was then segmented out by iterating up each pixel in each column until a threshold increase in pixel intensity. The pixel height between the RPE and ONL / OPL interface was then calculated for each column. Errors in ONL / OPL interface detection were reduced by excluding heights several standard deviations above the mean. The average height of the ONL / OPL interface above the RPE was subtracted from the height at every column to obtain the deviation of the ONL / OPL interface at every column above the average. Positive deviations were then summed to get the total volume of deviation of the OPL / ONL interface above average for each B-scan. This value was then averaged for all 49 B-scans in an OCT acquisition to calculate the RIPL index. Count the number of RIPLs in each OCT volume (B-scans) A further method utilizes a computer vision heuristic model to count the number of RIPLs in each OCT volume (B-scans). The same method was used as in the computer vision heuristic model to calculate a RIPL index to segment out the RPE and the OPL / ONL interface. A best fit spline was then used to better trace the RPE and filter out noise. That best fit spline was then increased from1 pixel in thickness to 7 pixels in thickness. The slice of 7 pixels in height was then moved up by a proportion of the average distance between the RPE and ONL / OPL interface to trace the OPL. The average pixel intensity for each column in the 7-pixel height RPE slice and 7-pixel height OPL slice were then calculated. Columns where the RPE slice average pixel intensity were a certain proportion below average were excluded 45667975.1 64 from the OPL slice to account for shadowing. Multiple consecutive dark columns in the OPL slice were counted as a single RIPL, excluding multiple consecutive dark columns above a certain threshold which was attributed to the fovea. Machine learning object detection model A further method utilizes a machine learning object detection model to detect RIPLs. OCT scans from 78 subjects (156 eyes) were collected and amounted to 7644 B-scan images. The B-scan images were 504 pixels in width and were 498 pixels in height and formatted as .jpg images. The data was then partitioned into training data and testing data with a ratio of 0.8 (80% used for training). Subjects were never partitioned into both training and testing data. They were partitioned to one or the other so that subjects in the testing set were never encountered in the training data. The dataset for training included 62 subjects, 124 eyes, and 6076 B-scan images. The dataset for testing included 16 subjects, 32 eyes, and 1568 B-scan images. The RIPLs were then labeled using the labelImg library by a trained grader (MS). 121 B-scans from the training dataset contained RIPLs.31 B-scans from the training dataset contained RIPLs. Training data containing RIPLs was duplicated to create an even distribution between B-scans containing RIPLs and those not containing RIPLs for training. The B-scans were then preprocessed to crop the image to focus on the layers of interest using part of the computer vision algorithm described in the computer vision heuristic model to calculate a RIPL index above. A 30-pixel height slice that traced the RPE was shifted up a certain amount to capture the ONL, OPL, and INL. That portion of the image was cropped. The cropped image was then sliced in three parts horizontally and then stacked vertically to create an image closer to the dimensions of a square which could then be fed into the object detection model. Labeled bounding boxes were moved accordingly with the preprocessing. SSD MobileNet V2 FPNLite 320x320 COCO17 TPU-8 pretrained model was used as the starting point for our model. The SSD MobileNet V2 FPNLite is a state-of-the-art object detection model that combines the strengths of the single-shot multibox detector (SSD) and MobileNet V2, a lightweight and efficient convolutional neural network (CNN) architecture. This model is particularly suitable for our task due to its fast inference time and high accuracy, which are crucial for processing large volumes of OCT images in a clinical setting. It is based on the MobileNet V2 architecture, which employs depth-wise separable convolutions and an inverted residual structure, enabling it to achieve high performance with a relatively low number of parameters. This reduces the computational complexity of the 45667975.1 65 model, making it ideal for deployment on edge devices and real-time applications. In addition, the model incorporates a feature pyramid network (FPN), which is a multi-scale feature extraction technique that enhances the detection of objects at different scales and aspect ratios. This is particularly beneficial for the task, as RIPLs can vary in size and shape, making them challenging to detect consistently. By integrating the FPN with the SSD detection framework, the model is able to identify RIPLs more effectively and with improved precision. The model was trained on a local machine running Ubuntu 20.04 with an NVIDIA T4004GB GPU using Python version 3.8, Tensorflow 2.11.1, CUDA version 12.0, and CUDNN version 8.8.1. The batch size was 32. The model was trained over 10 epochs. A low number of epochs was chosen to avoid overfitting because the images with RIPLs had been duplicated to even the distribution between non-RIPL and RIPL containing images. Segmentation of the outer plexiform layer A further method involves segmentation of the outer plexiform layer which follows the curvature of the retinal pigment epithelium (or alternatively choriocapillaris or internal limiting membrane). Following the curvature allows to avoid sudden deviations from RIPLs. Alternative methos is to construct a best fit line at the level of the outer plexiform layer. The area is increased to thickness ranging from 1 pixel to 30 pixel or in between). RIPLs are identified as dark spots in the line containing the outer plexiform layer. References 1. Brundel, et al., Nat Rev Dis Primers. Apr 72022;8(1):21. 2. Rahman, et al. Nat Rev Cardiol. Nov 2014;11(11):639-54. 3. Lloyd-Jones, et al. Circulation. Aug 312004;110(9):1042-6. 4. Heeringa, et al. European Heart Journal.2006;27(8):949-953. 5. Odutayo, et al. BMJ.2016;354:i4482. doi:10.1136 / bmj.i4482 6. Madhavan, et al. Nature Reviews Cardiology.2018 / 12 / 012018;15(12):744-756. 7. Lip, et al. Jama. May 192015;313(19):1950-62. 8. Kotalczyk, et al. Nat Rev Cardiol. Apr 2021;18(4):276-290. 9. Lip, et al. Chest. Feb 2010;137(2):263-72. 10. Ntaios, et al. Neurology. Mar 122013;80(11):1009-17. 45667975.1 66 11. Hindricks, et al. Eur Heart J. Feb 12021;42(5):373-498. 12. Cannon, et al., JAMA Network Open.2023;6(4):e239638-e239638. 13. Chen, et al. Stroke.2019;50(1):28-33. 14. Chen, et al. J Geriatr Cardiol. Sep 2013;10(3):258-66. 15. Christiansen, et al.2018 / 05 / 102018;18(1):91. 16. Grory, et al. Stroke.2021;52(9):2773-2781. 17. Madala, et al. Ophthalmol Retina. Sep 2022;6(9):865-867. 18. Bakhoum, et al. J Am Heart Assoc. Aug 152023;12(16):e028853. 19. Scharf, et al. Prog Retin Eye Res. Mar 2021;81:100884. 20. Sarraf, et al. JAMA Ophthalmol. Oct 2013;131(10):1275-87. 21. Bakhoum, et al. American Journal of Ophthalmology.2018 / 11 / 01 / 2018;195:143-153. 22. Long, et al. EClinicalMedicine. Mar 2021;33:100775. 23. Go, et al. Jama. May 92001;285(18):2370-5. 24. Kim, et al. Circ Cardiovasc Qual Outcomes. May 2011;4(3):313-20. 25. Lee, et al. PLoS One.2018;13(12):e0209687. Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of skill in the art to which the disclosure belongs. Publications cited herein and the materials for which they are cited are specifically incorporated by reference. Those skilled in the art will recognize or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the disclosure. Such equivalents are intended to be encompassed by the following claims. 45667975.1 67

Claims

CLAIMS We claim:

1. A method comprising: (i) determining the presence or absence of retinal ischemic perivascular lesions (RIPLs) in an eye of a subject; and (ii) selecting the subject for treatment for increased risk of stroke when the determining in step (i) identifies one or more RIPLs in an eye of the subject.

2. The method of claim 1, wherein the subject has, or has been identified as being at increased risk of having, atrial fibrillation (AF) and / or other cardiovascular disease (CVD).

3. The method of claim 1, wherein the subject does not have atrial fibrillation (AF) and / or other cardiovascular disease (CVD).

4. The method of any one of claims 1-3 further comprising, prior to step (ii), determining a risk factor (CHA₂DS₂-VASc) score for the subject, wherein the risk factors comprise one or more of Congestive heart failure, Hypertension, Age≥75 years, Diabetes mellitus, Prior Stroke or Transient Ischemic Attack (TIA) or Thromboembolism, Vascular disease, Age 65–74 years, and Sex category.

5. The method of claim 4, further comprising: (iii) administering anti-ischemia treatment and / or anti-stroke treatment to the subject selected in step (ii) if the CHA₂DS₂-VASc score is 1 or above.

6. The method of claim 1 further comprising: (iii) administering anti-ischemia treatment and / or anti-stroke treatment to a subject selected for treatment in step (ii).

7. The method of claim 5 or 6, wherein the treatment comprises anti-coagulation therapy.

8. The method of claim 7, wherein the anti-coagulation therapy comprises administering an anticoagulant agent to the subject, wherein the anticoagulant agent is selected from the group consisting of warfarin, apixaban, dabigatran, edoxaban, rivaroxaban, heparin, and aspirin.

9. The method of any one of claims 5 to 8, wherein the treatment comprises surgery.

10. The method of claim 1 further comprising excluding the subject from a treatment or refraining from treating the subject when the subject is not selected in step (ii).

11. The method of claim 4 further comprising excluding the subject from a treatment or refraining from treating the subject when the CHA₂DS₂-VASc score is below 1. 45667975.1 6812. The method of claim 1, further comprising (a) discontinuing anti-ischemia treatment and / or anti-stroke treatment; or (b) refraining from further anti-ischemia treatment and / or anti-stroke treatment of the subject that is or had been receiving anti-ischemia treatment and / or anti-stroke treatment.

13. The method of any one of claims 1 to 10, wherein the determining in step (i) comprises imaging one or both eyes of the subject.

14. The method of claim 13, wherein the imaging comprises optical coherence tomography (OCT).

15. The method of claim 14, wherein the OCT is spectral domain optical coherence tomography (SD-OCT).

16. The method of claim 14, wherein the OCT is time domain optical coherence tomography (TD-OCT).

17. The method of claim 14, wherein the OCT is swept source optical coherence tomography (SS-OCT).

18. The method of claim 14, wherein the OCT is visible light optical coherence tomography (VL-OCT).

19. The method of claim 12, wherein the OCT is macular optical coherence tomography (OCT).

20. The method of any one of claims 1 to 19, wherein the determining in step (i) comprises: (a) providing an OCT scan image of an eye of a subject; (b) identifying a cross section in the image corresponding to the retina which is flanked by the vitreous cavity and choroid; and (c) identifying the presence of RIPLs within the cross section in the image, optionally wherein the cross section comprises the inner nuclear layer, the outer plexiform layer and the outer nuclear layer of the eye.

21. The method of claim 20, wherein identifying the presence of RIPLs within the cross section in step (c) comprises: (I) fitting a polynomial curve to top points, bottom points, and A center curve passing through the midpoint of the cross section to provide a fitted image; (II) optionally processing the fitted image to remove noise to provide a processed image, wherein the processing optionally comprises one or more of thresholding, image dilation, and erosion; 45667975.1 69(III) filtering the fitted and / or processed image to accentuate horizontal lines using specified kernels to provide a filtered image, optionally, wherein the filtering comprises Median filtering and / or Kernel filtering; (IV) re-tracing lines and the center curve in the filtered image to provide a retraced image; and (V) identifying and / or quantifying RIPLs on the center lines and curve, optionally wherein the identifying comprises normalization of slopes compared to reference line.

22. The method of claim 20, wherein identifying the presence of RIPLs within the cross section in step (c) comprises: visual inspection of the cross section images.

23. The method of claim 22, wherein the presence of RIPLs is indicated by an observed expansion in the outer nuclear layer hypo-reflectivity.

24. The method of claim 22 or 23, wherein the visual inspection comprises duplicate, triplicate, or further inspections.

25. The method of claim 24, wherein the duplicate, triplicate, or further inspections are carried out by two or more different inspectors.

26. The method of any one of claims 20-25, further comprising annotating the identified RIPLs on the original image, and / or quantitating and / or annotating the total number of RIPLs per scan; and / or quantitating and / or recording a total number of RIPLs per eye and / or per subject.

27. The method of any one of claims 1 to 19, wherein the determining in step (i) comprises detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) via visual inspection of one or more image(s) of the inner retinal layer (IRL) of an eye of the subject, optionally wherein (a) the presence of RIPLs is indicated by an expansion in the outer nuclear layer hypo-reflectivity; and / or (b), wherein the visual inspection comprises duplicate, triplicate, or further inspections, optionally wherein the duplicate, triplicate, or further inspections are carried out by two or more different inspectors. 45667975.1 7028. The method of any one of claims 1 to 21, wherein the determining in step (i) comprises detecting the presence or absence of retinal ischemic perivascular lesions (RIPLs) in the eye of the subject via a computer-based system for the automated analysis of images of the inner nuclear layer, outer nuclear layer, inner plexiform layer, and outer plexiform layer of an eye of the subject.

29. The method of claim 28, wherein the computer-based system for the automated analysis of images comprises: (a) providing input data to a computer processor operably connected to means for computing and displaying output data, wherein the input data comprises one or more images of the inner nuclear layer, outer nuclear layer, inner plexiform layer and outer plexiform layer (retinal layer) of an eye of the subject, or data derived therefrom, wherein the output data comprises the number of RIPLs detected within the one or more images; and (b) computing the output data.

30. The method of claim 29, wherein the input data comprises a computer-readable media.

31. The method of any one of claims 1 to 19, wherein the detecting in step (i) comprises determining an RIPL index for one or more OCT B-scan image(s) of an eye of the subject, wherein the RIPL index quantifies total positive deviation of interface between an outer plexiform layer (OPL) and an outer nuclear layer (ONL) above an average position for the one or more B-scan image(s).

32. The method of claim 31, wherein determining the total positive deviation of the OPL / ONL interface above average comprises: (a) providing one or more OCT B-scan image(s) of an eye of a subject, wherein a scan image comprises a multiplicity of pixel intensities, wherein a pixel intensity of the multiplicity of pixel intensities comprises a value, and wherein the multiplicity of pixel intensity values are disposed as columns and rows throughout the image; (b) processing the OCT B-scan image(s) to reduce noise and highlight internal limiting membrane (ILM) and retinal pigment epithelium (RPE) layers in the image(s), to provide a processed scan having columns of pixel intensity, 45667975.1 71optionally wherein the highlighting comprises one or more processing steps of erosion, dilation, and / or thresholding; (c) segmenting the ILM and RPE, wherein the segmenting comprises iterating down each column of pixel intensity values and up each column of pixel intensity values, respectively, until there is a change in pixel intensity value of at least a first threshold amount; (d) excluding scans of fovea, wherein the excluding comprises determining an average central retinal thickness, and removing data from OCT B-scan image(s) having a central retinal thickness that is a second threshold amount less than the average retinal thickness; (e) locating the darkest pixel in each column to identify the outer plexiform layer (OPL), wherein the darkest pixel has the greatest intensity value; (f) segmenting out the outer nuclear layer (ONL) / OPL interface, wherein the segmenting comprises iterating up each pixel in each column until a third threshold increase in pixel intensity; (g) calculating the pixel height between the RPE and ONL / OPL interface for each column; (h) optionally excluding heights several standard deviations above the mean to reduce errors in ONL / OPL interface detection; (i) obtaining positive deviations of the ONL / OPL interface, wherein the obtaining comprises determining the average height of the ONL / OPL interface above the RPE, and subtracting the average height of the ONL / OPL interface above the RPE from the height at every column above the average; and (j) summing the positive deviations to provide the total volume of deviation of the OPL / ONL interface above average for the one or more OCT B-scan image(s).

33. The method of claim 32, wherein step (b) further comprises removing shadowing from the images, wherein the removing comprises providing a best fit spline having a single pixel thickness to trace the RPE; tracing the OPL, wherein tracing the OPL comprises increasing the thickness of the best fit spline from 1 pixel in thickness to up to about 7 pixels in thickness and moving the spline up by a proportion of the average distance between the RPE and ONL / OPL interface; calculating the average pixel intensity for each column in the 7-pixel height RPE slice and 7-pixel height OPL slice; and 45667975.1 72excluding a column from the OPL slice to account for shadowing when the RPE slice average pixel intensity are a fourth threshold value below average.

34. The method of claim 33, further comprising: identifying and counting multiple consecutive columns in the OPL slice having pixel intensity above the fourth threshold value as a single RIPL; and excluding multiple consecutive dark columns above a fifth threshold value as being attributed to fovea.

35. The method of any one of claims 1 to 19, wherein RIPLs are identified from data within one or more OCT B-scan scan of an eye of the subject, wherein the identification comprises utilizing a computer program run on a computer system, wherein the program is configured to identify RIPLs in the OCT B-scan by segmentation of the outer plexiform layer (OPL) which follows the curvature of the retinal pigment epithelium.

36. The method of any one of claims 31 to 35, wherein the method is repeated one or more times to sequentially identify RIPLs in scans from two or more subjects or pools of subjects.

37. The method of claim 36, wherein the repeated method is carried out sequentially by a computer as a first and second or further iterations, wherein the computer implements a machine-learning system to increase the efficacy of the method in a second or further iteration as compared with a first or further iteration.

38. The method of any one of claims 1 to 37, wherein the method further comprises, prior to step (ii), one or more steps to determine whether the subject has AF.

39. The method of any one of claims 1 to 38, wherein the AF is asymptomatic AF or silent AF.

40. The method of any one of claims 1 to 39, wherein the subject has a personal or family history of diseases or disorder selected from the group consisting of AF, cardiovascular disease, hypertension, carotid disease, diabetes, hypercholesterolemia, hyperlipidemia, stroke, and ischemic disease.

41. The method of any one of claims 1 to 40, wherein the subject has an implantable cardiac device, and / or has had recent cryptogenic stroke or transient ischemic attack (TIA), or vascular dementia. 45667975.1 7342. The method of any one of claims 1 to 41, wherein the subject has a personal or family history of diseases or disorder selected from the group consisting of retinopathy, diabetic retinopathy, hypertensive retinopathy, ischemic retinopathy, or macular edema. 45667975.1 74

Images