Diabetic retinopathy prediction method based on choroidal retina characteristic ratio

Abstract

The invention belongs to the field of intelligent medical treatment, and particularly relates to a diabetic retinopathy prediction method based on a choroidal retina characteristic ratio. The method comprises the following steps: acquiring an eye OCTA image of a diabetic patient; extracting choroid features and retina features based on the OCTA image, and calculating a choroid to retina feature ratio based on a ratio of the choroid features to the retina features; and inputting the choroid and retina feature ratio into a classifier to obtain a result whether the diabetic suffers from retinopathy or not. The application proposes and verifies the hypothesis that the retina may have the corresponding retinal blood supply system reserve (RBSSR), and extracts the choroid-retina feature ratio with important prediction performance for the monitoring of the diabetic retinopathy based on the hypothesis, thereby achieving the timely and accurate prediction effect of the diabetic retinopathy.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medicine, and more specifically, to a prediction method, device, medium and program product for diabetic retinopathy based on the choroid-retina feature ratio. Background Art

[0002] Diabetic retinopathy (DR) is one of the most common and severe microvascular complications of diabetes mellitus (DM), and has become the main cause of blindness in the working-age population worldwide. However, the pathogenesis of DR remains unclear. Although the duration of diabetes and blood glucose control are widely regarded as key risk factors for DR, clinical observations have found that: some patients with good blood glucose control and short duration of diabetes still develop DR, while some patients with poor blood glucose control and long duration of diabetes still do not have DR. This difference indicates that individual differences play an important role in the occurrence and development of DR.

[0003] The retina is one of the tissues with the highest oxygen consumption in the human body. In patients with DR, retinal ischemia and hypoxia caused by impaired microcirculation can lead to neurodegeneration, retinal dysfunction, and even vision loss. The concept of cerebrovascular reserve refers to the ability of the cerebrovascular neurovascular unit to dynamically regulate blood pressure, intracranial pressure fluctuations, and increased metabolic demands. This adaptability is usually used as a key indicator to evaluate the occurrence and progression of cerebrovascular diseases including diabetes.

[0004] In the prior art, "Assessment of choroidal structural changes in patients with pre- and early-stage clinical diabetic retinopathy using wide-field SS-OCTA" discloses the use of SS-OCTA to study the microvascular changes in the choroidal structure of patients with pre-clinical and early-stage diabetic retinopathy (DR). By recording and analyzing the choroidal vascularity index (CVI), choroidal thickness (ChT), and central macular thickness (CMT) of the entire area (12 mm in diameter) and concentric rings in different ranges (0-3, 3-6, 6-9, and 9-12 mm), it is found that the CVI and ChT values of diabetic patients are significantly lower than those of healthy controls, especially in patients with early DR. In addition, the peripheral choroidal capillaries are more vulnerable to early DM-induced damage than the central region. However, the role of the retinal vascular supply system composed of the retinal circulation (RC) and choroidal circulation (CC) in DR is not disclosed in the prior art. Summary of the Invention

[0005] In view of the above problems, the present invention provides a prediction method for diabetic retinopathy based on the choroid-retina feature ratio, which realizes the prediction of diabetic retinopathy by using the relative changes in the volumes of the choroid and the retina.

[0006] The present application (in the first aspect) discloses a prediction method for diabetic retinopathy based on the choroid-retina feature ratio, including: Obtain the ocular OCTA images of diabetic patients; Extract choroid features and retina features based on the OCTA images, and calculate the choroid-retina feature ratio based on the ratio of the choroid features and the retina features; Input the choroid-retina feature ratio into a classifier to obtain the result of whether diabetic patients have retinopathy.

[0007] Furthermore, the choroid-retina feature ratio includes any one or several of the following: The ratio of the total choroid volume to the total retina volume, the ratio of the choroid vascular volume to the total retina volume, The ratio of the total choroid volume to the outer retina volume, the ratio of the choroid vascular volume to the outer retina volume.

[0008] Furthermore, extract the choroid features and retina features of the central region where the fovea is located based on the OCTA images, and calculate the choroid-retina feature ratio based on the ratio of the choroid features and the retina features of the central region.

[0009] Furthermore, the choroid feature is the total choroid volume, the retina feature is the total retina volume, and the choroid-retina feature ratio is the ratio of the total choroid volume to the total retina volume; Furthermore, the ratio of the total choroid volume to the total retina volume is the ratio of the total choroid volume to the total retina volume of the central region; Furthermore, the choroid feature is the total choroid volume, the retina feature is the total retina volume, and the choroid-retina feature ratio is the ratio of the total choroid volume to the total retina volume; Furthermore, the ratio of the total choroid volume to the total retina volume is the ratio of the total choroid volume to the total retina volume of the central region; Input the ratio of the total choroid volume to the total retina volume of the central region into a classifier to obtain the result of whether diabetic patients have retinopathy.

[0010] Furthermore, the choroid feature is the choroid thickness, the retina feature is the retina layer thickness, and the choroid-retina feature ratio is the ratio of the choroid thickness to the retina layer thickness; Further, the choroid-to-retina feature ratio is the ratio of the choroid thickness to the retina layer thickness in the central region; Input the ratio of the choroid thickness to the retina layer thickness in the central region into the classifier to obtain the result of whether a diabetic patient has retinopathy.

[0011] Further, the choroid feature is the choroid thickness, the retina feature is the outer retina thickness, and the choroid-to-retina feature ratio is the ratio of the choroid thickness to the outer retina thickness; Further, the ratio of the total choroid volume to the total retina volume is the ratio of the choroid thickness to the outer retina thickness in the central region; Input the ratio of the choroid thickness to the outer retina thickness in the central region into the classifier to obtain the result of whether a diabetic patient has retinopathy.

[0012] Further, the choroid feature is the total choroid volume, the retina feature is the outer retina volume, and the choroid-to-retina feature ratio is the ratio of the total choroid volume to the outer retina volume; Further, the ratio of the total choroid volume to the outer retina volume is the ratio of the total choroid volume to the outer retina volume in the central region; Input the ratio of the total choroid volume to the outer retina volume in the central region into the classifier to obtain the result of whether a diabetic patient has retinopathy.

[0013] Further, the choroid feature is the choroid vascular volume, the retina feature is the total retina volume, and the choroid-to-retina feature ratio is the ratio of the choroid vascular volume to the total retina volume; Further, the ratio of the choroid vascular volume to the total retina volume is the ratio of the total choroid volume to the total retina volume in the central region; Input the ratio of the choroid vascular volume to the total retina volume in the central region into the classifier to obtain the result of whether a diabetic patient has retinopathy.

[0014] Further, the choroid feature is the choroid vascular volume, the retina feature is the outer retina volume, and the choroid-to-retina feature ratio is the ratio of the choroid vascular volume to the outer retina volume; Further, the ratio of the total choroid volume to the total retina volume is the ratio of the total choroid volume to the total retina volume in the central region; Input the ratio of the choroid vascular volume to the outer retina volume in the central region into the classifier to obtain the result of whether a diabetic patient has retinopathy.

[0015] Further, the total choroid volume is the volume between the RPE / BM in the central region and the choroid-scleral interface, The choroidal vascular volume is the volume of large and medium-sized blood vessels in the choroid within the central region; The total retinal volume is the volume between the inner limiting membrane and the RPE / BM within the central region; The outer retinal volume is the volume between the outer plexiform layer and the RPE / BM within the central region.

[0016] Further, the OCTA image is divided into a central region where the fovea is located and a peripheral region outside the fovea, and the choroidal features, retinal features in the central region, and choroidal features, retinal features in the peripheral region are extracted respectively; the choroid-to-retina feature ratio is calculated based on the ratio of the choroidal features and retinal features in the central region, and the choroid-to-retina feature ratio in the peripheral region is calculated based on the ratio of the choroidal features and retinal features in the peripheral region.

[0017] Further, the choroid-to-retina feature ratio further includes any one or several of the following: The ratio of the total choroidal volume in the peripheral region to the total retinal volume, the ratio of the choroidal vascular volume in the peripheral region to the total retinal volume, The ratio of the total choroidal volume in the peripheral region to the outer retinal volume, the ratio of the choroidal vascular volume in the peripheral region to the outer retinal volume.

[0018] Further, the ocular OCTA imaging is obtained by performing OCTA imaging in a scanning mode centered on the fovea, and the obtained image is evenly divided into a 3×3 grid. The central grid is designated as the central region, and the surrounding grids are regarded as the peripheral region.

[0019] Further, concentric circles with different radii are set with the fovea center as the center of the circle. The region within a radius of 6 mm is the central region, and the region from 6 mm to 12 mm is the peripheral region.

[0020] Further, the construction method of the classifier is as follows: Obtain the ocular OCTA images of diabetic patients in the training set and the corresponding labels, and the labels include no retinopathy and retinopathy; Extract the choroidal features and retinal features from the ocular OCTA images in the training set, and calculate the choroid-to-retina feature ratio; Input the choroid-to-retina feature ratio and the labels into a machine learning model for iterative training to obtain the classifier.

[0021] The second aspect of the present application discloses a prediction system for diabetic retinopathy based on the choroid-to-retina feature ratio, including: An acquisition module 201: used to acquire the ocular OCTA images of diabetic patients; Feature extraction module 202: used to extract choroid features and retina features based on the image, and calculate the choroid-to-retina feature ratio based on the ratio of the choroid features and retina features; Prediction module 203: used to input the choroid-to-retina feature ratio into a classifier to obtain the result of whether a diabetic patient has retinopathy.

[0022] The third aspect of the present application discloses a computer device, the device includes: a memory and a processor; the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, it is used to execute the steps of the above method.

[0023] The fourth aspect of the present application discloses a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the above method.

[0024] The fifth aspect of the present application discloses a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the steps of the above method.

[0025] The present application has the following beneficial effects: (1) The present application proposes and verifies the hypothesis that the retina may have a corresponding retinal blood supply system reserve (RBSSR), and extracts the choroid-retina feature ratio with important predictive performance for the detection of diabetic retinopathy based on this hypothesis, realizing the prediction effect of diabetic retinopathy; (2) The clinical index of the choroid-to-retina volume ratio proposed in the present application has high prediction sensitivity for diabetic retinopathy, and can detect the risk of retinopathy in diabetic patients in a timely manner. Description of the Drawings

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0027] Figure 1 It is a schematic flowchart of the method provided in the first aspect of the embodiments of the present invention; Figure 2 It is a schematic diagram of the program product provided in the second aspect of the embodiments of the present invention; Figure 3 It is a schematic diagram of the computer device provided in the embodiments of the present invention; Figure 4 It is a schematic diagram of the architecture of an exemplary computing device provided in the embodiments of the present invention; Figure 5 Schematic diagram of the storage medium provided by an embodiment of the present invention; Figure 6 It is a schematic diagram of OCTA image acquisition and analysis provided by an embodiment of the present invention. Among them, a 12 × 12 mm 2 scanning mode is adopted to take OCTA fundus images centered on the macula, and the built-in software is used for automatic stratification and analysis: the retinal layer refers to the area from the inner limiting membrane (ILM) to the retinal pigment epithelium / Bruch's membrane (RPE / BM); the outer retinal layer refers to the area from the outer plexiform layer (OPL) to the RPE / BM; the choroidal layer refers to the area from the RPE / BM to the choroidal-scleral interface (CSI). (A-B) The OCTA scan results of each retinal and choroidal layer are divided into 3 × 3 grids of 4 × 4 mm 2 (the central 4×4 mm grid is defined as the central region, and the remaining grids are defined as the peripheral regions), and then the relevant retinal and choroidal volumes are measured. (C-D) Schematic diagram of the B-scan mode of the volume ratio of the choroid to the retina: TCV / ORV, defined as the ratio of the total choroidal volume to the outer retinal volume; TCV / TRV, defined as the ratio of the total choroidal volume to the total retinal volume; LCV / ORV, defined as the ratio of the choroidal vascular volume (LCV, yellow area) to the outer retinal volume; LCV / TRV, defined as the ratio of LCV to the total retinal volume; Figure 7 Schematic diagram of the increase in the volume ratio of the choroid to the retina in mice by atropine provided by an embodiment of the present invention; (A) shows the comparison of OCTA imaging before drug administration / control group and after drug administration, and compares the ratio of choroidal thickness between the control group and the experimental group; (B) is a comparison diagram of the retinal volume, choroidal volume, and choroid-retina volume ratio between the control group and the atropine experimental group. Figure 8 Schematic diagram of the correlation between low choroid-retina volume ratio and DR in DM mice provided by an embodiment of the present invention. Among them, (A) is a schematic diagram of the animal modeling experimental process, (B) is the wide-angle color fundus image (CF), fundus autofluorescence image (AF), and OCTA 6 X6 mm2 scan image of DM mice. The yellow arrow indicates the temporal side, n = 7; (C) Representative HE staining of the retinal tissue of DM mice, n = 6; (D) TUNEL staining of the retinal tissue of DM mice to detect apoptosis, n = 6; Paired t-tests were used for data analysis, and the results are expressed as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001. Detailed implementation manners

[0028] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0029] In some processes described in the specification, claims and above-mentioned drawings of the present invention, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The serial numbers of the operations, such as S101, S102, etc., are only used to distinguish each different operation, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent a sequence, and do not limit that "first" and "second" are different types.

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work belong to the scope of protection of the present invention.

[0031] Figure 1 It is a schematic flow chart of a prediction method for diabetic retinopathy based on evaluating the choroid-retina feature ratio provided by an embodiment of the present invention. Specifically, the method includes the following steps: S101: Obtain the ocular OCTA image of a diabetic patient; S102: Extract the choroid feature and retina feature based on the OCTA image, and calculate the choroid-retina feature ratio based on the ratio of the choroid feature and retina feature; S103: Input the choroid-retina feature ratio into a classifier to obtain the result of whether the diabetic patient has retinopathy.

[0032] Given that the retina is part of the central nervous system, we have reason to assume that the retina may also have a corresponding retinal blood supply system reserve (RBSSR), that is, the ability of the retinal vascular supply system to meet increased metabolic demands. This reserve capacity may play a key role in the pathogenesis of DR.

[0033] In the human body, the retinal vascular supply system consists of the retinal circulation (RC) and the choroidal circulation (CC). RC accounts for only a small part of the retinal blood supply, while CC accounts for approximately 85% of the retinal blood supply, mainly supplying the outer five layers of the retina and the macula. In addition, RC is characterized by low blood flow and high oxygen uptake rate, resulting in a large arteriovenous oxygen saturation difference and an extremely sensitive response to hypoxia. In contrast, CC has a large blood flow but a very low oxygen uptake rate, resulting in a small arteriovenous difference and a high reserve of oxygen transport capacity, indicating strong tolerance to hypoxia. Given these characteristics, CC may be the main contributor to RBSSR.

[0034] Since the choroid is mainly composed of vascular tissue, theoretically its volume is positively correlated with the oxygen transport capacity of CC. As the retina is the most oxygen-consuming tissue in the human body, its size directly affects its oxygen demand.

[0035] Based on the above theoretical analysis, it is concluded that the choroid-to-retina volume ratio can theoretically serve as a relative indicator of RBSSR. A detailed examination of this ratio may help evaluate RBSSR and deepen our understanding of the potential pathogenesis of DR. Combining the above theoretical insights and clinical observations, we hypothesize that individuals with relatively low RBSSR may be more prone to developing DR. Specifically, this hypothesis suggests that DR patients usually exhibit a lower choroid-to-retina volume ratio. To test this hypothesis, we used ultra-widefield swept-source optical coherence tomography angiography (OCTA) to analyze the retinal and choroidal vascular images of DM patients and found that a low choroid-to-retina volume ratio is a risk factor for DR.

[0036] Previous reports have shown that atropine can induce choroidal thickening in myopic patients or animals. Therefore, we established an animal model using atropine and further demonstrated that atropine treatment effectively alleviated early DR lesions, and this effect was related to an increase in the choroid-to-retina volume ratio.

[0037] Among them, the specific experimental studies are described as follows.

[0038] I. Materials and Methods: 1.1 Research Subjects This is a cross-sectional study, and the research subjects are type 2 diabetes mellitus (T2DM) patients recruited from the Ophthalmology Department of the Third Affiliated Hospital of Southern Medical University in Guangzhou, China, from June 1, 2022, to June 1, 2024. According to the International Clinical Diabetic Retinopathy Disease Severity Scale, the subjects were divided into two groups. The control group was T2DM patients without obvious retinopathy (disease duration ≥ 5 years) (NDR group, 79 cases / 134 eyes), and the DR group was T2DM patients with non-proliferative DR (disease duration not limited) (89 cases / 125 eyes).

[0039] All subjects underwent ophthalmic examinations, including best-corrected visual acuity (BCVA), intraocular pressure (IOP), slit-lamp fundus examination, fundus photography, axial length (AL), spherical equivalent (SE), OCT, and OCTA. Demographic data of all participants were collected, including age, gender, body mass index (BMI), fasting blood glucose (FBG) level, duration of diabetes, history of hypertension, and status of chronic kidney disease.

[0040] The inclusion criteria for the subjects were as follows: (1) age ≥ 18 years; (2) diagnosis of T2DM; (3) duration of T2DM ≥ 5 years in the NDR group. The exclusion criteria included: (1) intraocular pressure < 21 mmHg; (2) blood pressure ≥ 160 / 100 mmHg; (3) presence of diabetic macular edema, epiretinal membrane, or pathologic myopia; (4) history of other intraocular surgeries other than cataract, including retinal photocoagulation; (5) poor quality of OCTA images due to opaque refractive media, etc., with a signal score < 7 (on a 10-point scale); (6) other ocular diseases that may affect ocular circulation, such as glaucoma, age-related macular degeneration, retinal vascular diseases, etc.

[0041] 1.2 Experimental animals Three-week-old male C57BL / 6J mice were purchased from Guangzhou Regene Biotech Co., Ltd. Atropine eye drops at a concentration of 2% were given to one eye twice a day, while the contralateral eye received saline eye drops as a control. Diabetes was induced by intraperitoneal injection of streptozotocin (STZ) at 55 mg / kg, dissolved in sodium citrate buffer (pH 4.5), for 5 consecutive days. Blood glucose was measured 1 week after induction, and a blood glucose level ≥ 16.7 mmol / L confirmed diabetes.

[0042] 1.3 OCTA image acquisition and analysis All OCTA images were taken using a BM-400K (BMizar, TowardPi Medical Technology Co., Ltd., Beijing, China) at the same time period of the day (8:00 - 17:00). For human subjects, OCTA imaging was performed using a 12 × 12 mm 2 scanning pattern centered on the fovea ( Figure 6 A). The obtained images were divided into a 3 × 3 grid of 4 × 4 mm 2 .

[0043] The central grid was designated as the central area, while the surrounding grids were regarded as the peripheral area ( Figure 6 B).

[0044] 1.3.1 Volume measurement The volumes of the retina and choroid were measured after automatic stratification using the built-in software: AsFigure 6 As shown in D, the retinal layer refers to the region from the inner limiting membrane (ILM, marked by the red line) to the retinal pigment epithelium / Bruch's membrane (RPE / BM); the total retina volume (TRV) corresponds to Figure 6 the volume of the part marked by the blue line segment in D; the total retina volume includes the total volumes of the inner layer (from ILM to INL) and the outer layer (from ONL to RPE / BM), that is, the total volume from ILM to RPE / BM. The outer retina (or outer layer) refers to the region from the outer plexiform layer (OPL) to RPE / BM; the outer retinal volume (ORV) corresponds to the volume of the part marked by the red line segment in Figure 6D.

[0045] The outer retina (or outer layers) refers to the region from the outer plexiform layer (OPL) to RPE / BM; the outer retinal volume (ORV) corresponds to Figure 6 the volume of the part marked by the magenta line segment in D; The choroidal layer refers to the layer from RPE / BM to the choroid-scleral interface (CSI), and the total choroidal volume (TCV) corresponds to Figure 6 the total volume of the region marked by the white line segment in D; The choroidal layer is an important part of the middle membrane structure of the eyeball wall, located between the outer retina and the sclera, mainly composed of a vascular layer, connective tissue, and melanocytes. The choroidal vascular system is the most core anatomical structure of this layer. The choroidal vascular system is supplied by the short posterior ciliary arteries and forms a unique layered distribution, including: the Haller layer (large vessel layer), the Sattler layer (medium vessel layer), and the choriocapillaris; in this study, the volume of the large and medium choroidal vessels was measured as the choroidal vascular volume (LCV), as Figure 6 shown by the yellow block in D, which marks the large vessel layer and the medium vessel layer in the choroid.

[0046] 1.3.2 Volume calculation Based on the above volume measurements, the following ratios are calculated: TCV / ORV, defined as the ratio of the total choroidal volume to the outer retinal volume; TCV / TRV, defined as the ratio of the total choroidal volume to the total retina volume; LCV / ORV, defined as the ratio of the choroidal vascular volume (LCV) of the LMCV to the outer retina volume; LCV / TRV, defined as the ratio of LCV to the total retina volume; Treatment of experimental mice: In the OCTA 6x6 mm 2 mode, fundus images of mice centered on the ONH. Then the volumes of the retina and choroid were automatically measured by the built-in software. Mouse OCTA 6x6 mm 2 scanning images. Slices with the scanning plane horizontally passing through the center of the ONH were selected for choroidal thickness measurement ( Figure 7 A).

[0047] 1.4 Statistical analysis Statistical analysis was performed using GraphPad Prism version 9.0.0. The Shapiro-Wilk test was used to evaluate the normality of data distribution. According to the distribution, the data were expressed as mean ± standard deviation (SD) or median (interquartile range, IQR). Categorical data were expressed as percentages. The BCVA values were transformed into the logarithm of the minimum angle of resolution (logMAR) for analysis. For continuous data comparison, independent sample t-test, Welch's t-test or Mann-Whitney U test were used as appropriate. Categorical data comparison was performed using chi-square test or Fisher's exact test. Univariate logistic regression analysis was used to evaluate the correlation. Specifically, paired t-test was used to analyze the animal experiment data. The P values were two-sided, and P < 0.05 was considered statistically significant.

[0048] 1.5 Fundus photography and autofluorescence images After induction of anesthesia in mice, wide-angle color fundus imaging and autofluorescence images centered on the ONH were obtained using Optos 200 Tx (Optos PLC, Dunfermline, United Kingdom).

[0049] 1.6 H&E staining and TUNEL assay The mice were euthanized, the eyeballs were removed, and fixed with FAS eye fixative (Servicebio) for 24 hours. 4-μm tissue sections were prepared. Subsequently, hematoxylin and eosin (H&E) staining was performed using an H&E staining kit (Solarbio) according to the manufacturer's instructions.

[0050] 2 Experimental results 2.1 Demographic and clinical characteristics This study included 134 eyes of 79 NDR patients and 125 eyes of 89 DR patients. There were no significant differences in age, gender, diabetes duration, FBG, hypertension, BMI, BCVA, IOP, AL, and SE between the two groups (P < 0.05). The proportion of patients with chronic kidney disease was lower in the NDR group than in the DR group (P = 0.037). Table 1 provides a summary of the baseline characteristics of each group (Table 1).

[0051] Table 1. Demographic and Clinical Characteristics of the Subjects

[0052]

[0053] Table Note: * indicates P < 0.05; BMI = body mass index; HTN = hypertension; CKD = chronic kidney disease; FBG = fasting blood glucose; BCVA = best corrected visual acuity; IOP = intraocular pressure; SE = spherical equivalent; AL = axial length; logMAR = logarithm of the minimum angle of resolution visual acuity; SD = standard deviation; IQR = interquartile range.

[0054] 2.2 Choroid-Retina Volume Ratio Analysis

[0055] Compared with the NDR group, the volume in both the central and peripheral regions of the choroid was significantly reduced in the DR group (P < 0.001). The central region LCV in the DR group (1.601 ± 0.618 vs. 1.154 ± 0.674, P < 0.001) and the peripheral region LCV (9.600 ± 3.266 vs. 7.136 ± 3.203, P < 0.001) were significantly lower than those in the NDR group. There were no significant differences in the volume of the outer retinal layer and the total retinal volume between the two groups in any region (P < 0.05). (Table 2).

[0056] Analysis of the TCV / ORV, TCV / TRV, LCV / ORV, and LCV / TRV ratios showed that these ratio values were significantly lower in DR patients compared with the NDR group. This difference was consistent in both the central and peripheral regions, and all comparisons were statistically significant (P < 0.001, Table 2).

[0057] Table 2. Retinal and Choroidal Volumes and Choroid-to-Retina Volume Ratios

[0058] Table note: * indicates P < 0.05; Abbreviations and acronyms: LMCV (large and medium choroidal vessels): large and medium choroidal vessels; LCV (choroidal vascularity volume of LCMV): choroidal vascularity volume of large and medium choroidal vessels; TCV / ORV = ratio of total choroidal volume to outer retinal volume; TCV / TRV = ratio of total choroidal volume to total retinal volume; LCV / ORV = ratio of large and medium choroidal vessels volume to outer retinal volume; LCV / TRV = ratio of large and medium choroidal vessels volume to total retinal volume; SD = standard deviation.

[0059] 2.3 Logistic regression analysis

[0060] Univariate logistic regression analysis showed that low choroidal / retinal volume ratios, including TCV / ORV, TCV / TRV, LCV / ORV, and LCV / TRV were significantly associated with DR. Notably, the LCV / TRV ratio in the central region (OR 0.0019, 95% CI 0.0001 - 0.0181, P < 0.001) and the LCV / TRV ratio in the peripheral region (OR 1.254×10-4, 95%CI 4.969×10-6 - 2.469×10-3, P < 0.001) were both strongly negatively associated with DR (Table 3).

[0061] Table 3. Univariate Logistic regression analysis of the correlation between choroid-retina volume ratio and DR

[0062] Table note: * P <0.05; Abbreviations and acronyms: TCV / ORV = ratio of total choroidal volume to outer retinal volume; TCV / TRV = ratio of total choroid volume to total retina volume; LCV / ORV = ratio of LCV to outer retinal volume; LCV / TRV = ratio of LCV to total retina volume.

[0063] In some embodiments, the occurrence of DR is predicted based on the LCV / TRV ratio (OR 0.0019, 95% CI 0.0001 ~ 0.0181, P < 0.001). If the LCV / TRV ratio is higher than the threshold, the occurrence of DR is predicted; otherwise, it is predicted that DR does not occur, where the threshold is determined based on the above statistical analysis of the training set data.

[0064] In some embodiments, the occurrence of DR is predicted based on the LCV / TRV ratio in the surrounding area (OR 1.254×10-4, 95% CI 4.969×10-6 ~ 2.469×10-3, P < 0.001). If the LCV / TRV ratio is higher than the threshold, the occurrence of DR is predicted; otherwise, it is predicted that DR does not occur, where the threshold is determined based on the above statistical analysis of the training set data.

[0065] Based on the results of univariate Logistic regression analysis, we established a prediction model for DR using the features with significantly different inter-group differences found: The eigenvalue and corresponding label of the NDR group and DR group in the study were extracted (the NDR label is 0, and the DR label is 1), a data set was constructed, and the model was trained with 70% of the data as the training set and 30% of the data as the test set according to the random division method. The performance of the model is shown in Table 4: Table 4 Receiver operating characteristic (ROC) curve analysis of choroid-to-retina volume ratio for differentiating non-diabetic retinopathy (NDR) and diabetic retinopathy (DR)

[0066]

[0067] Table note: * indicates P < 0.05; abbreviations and acronyms: TCV / ORV = ratio of total choroid volume to outer retinal volume; TCV / TRV = ratio of total choroid volume to total retinal volume; LCV / ORV = ratio of LCV (local choroid volume) to outer retinal volume; LCV / TRV = ratio of LCV to total retinal volume; AUC = area under the ROC curve.

[0068] 2.4 Atropine increases the choroid-to-retina volume ratio in mice To further verify the significance of the choroid-to-retina volume ratio in DR, we conducted animal experiments. Initially, we explored the feasibility of using atropine to increase the choroid-to-retina volume ratio in mice. The OCT results showed that the choroid (temporal and nasal) of the eyes treated with atropine was significantly thicker than that of the control eyes ( Figure 7 A). At the same time, the OCTA results showed that we successfully established a mouse model with a significant difference in the choroid-to-retina volume ratio between the two eyes ( Figure 7 B).

[0069] Atropine (2%) was instilled into one eye of 3w mice, and saline was instilled into the contralateral eye twice a day for 6 weeks (n = 11). (A) OCTA 6X6 mm 2 scanning images of mice. Data before and after instillation were collected, and the changes in choroidal thickness at 2000 μm near the ONH were evaluated using the built-in software. The direction of the yellow arrow indicates the temporal side. (B) OCTA analysis of the changes in retinal and choroidal volumes, as well as the choroid-to-retina volume ratio. Data analysis was performed using paired t-tests, and the results are expressed as mean ± SD. * p < 0.05, ** p < 0.01.

[0070] 2.5. Low choroid-to-retina volume ratio is associated with DR in DM mice While maintaining atropine or saline eye drops, STZ was intraperitoneally injected to induce diabetes. Color fundus photography showed no significant difference between the two eyes. However, fundus autofluorescence imaging showed that there were fewer punctate hyperfluorescent lesions in the experimental eyes compared with the control group, indicating that atropine reduced the number of abnormal RPE cells in DM mice. In addition, OCTA showed that the choroid in the eyes treated with atropine was thicker than that in the control group ( Figure 8 B). HE staining showed that the outer nuclear layer (ONL) of DM mice was arranged loosely and disorderly, and this condition was alleviated after atropine treatment ( Figure 8 C).

[0071] Increased apoptosis of retinal cells is a key feature of early DR. Here, we demonstrated that atropine could reduce apoptosis in the outer retina ONL and RPE of DM mice ( Figure 8 D). Collectively, these data suggest that atropine can inhibit the progression of DR in DM mice, which is associated with an increased choroid-to-retina volume ratio. Therefore, it further corroborates that the choroid-to-retina volume ratio is a key clinical feature of diabetic retinopathy.

[0072] Figure 3 is a schematic diagram of a computer device provided by an embodiment of the present invention, as Figure 3 shown, the device 2000 may include: one or more processors 2010, and one or more memories 2020; wherein, computer-readable code is stored in the memory, and when the computer-readable code is run by the one or more processors, the method described above can be executed.

[0073] The processor in this embodiment may be an integrated circuit chip with signal processing capabilities. The above-mentioned processor may be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, operations and logic block diagrams disclosed in the embodiments of the present disclosure. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc., and may be of the X86 architecture or the ARM architecture.

[0074] Generally speaking, the various exemplary embodiments of the present disclosure may be implemented in hardware or dedicated circuits, software, firmware, logic, or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software that can be executed by a controller, a microprocessor or other computing devices. When aspects of the embodiments of the present disclosure are illustrated or described as block diagrams, flowcharts or using some other graphical representation, it will be understood that the blocks, devices, systems, techniques or methods described herein may be implemented as non-limiting examples in hardware, software, firmware, dedicated circuits or logic, general hardware or controllers or other computing devices, or some combination thereof.

[0075] For example, the method or device according to the embodiments of the present disclosure may also be implemented by means of Figure 4 the architecture of the computing device 3000 shown. As Figure 4 shown, the computing device 3000 may include a bus 3010, one or more CPUs 3020, a read-only memory (ROM) 3030, a random access memory (RAM) 3040, a communication port 3050 connected to a network, an input / output component 3060, a hard disk 3070, etc. The storage device in the computing device 3000, such as the ROM 3030 or the hard disk 3070, may store various data or files used for the processing and / or communication of the method provided by the present disclosure and the program instructions executed by the CPU. The computing device 3000 may also include a user interface 3080. Of course, Figure 4 the architecture shown is only exemplary, and when implementing different devices, one or more components shown in the Figure 4 computing device may be omitted according to actual needs.

[0076] The embodiment of the present invention also provides a computer-readable storage medium, such as Figure 5As shown, it is a schematic diagram of a storage medium 4000 provided by an embodiment of the present invention. Computer-readable instructions 4010 are stored on the computer storage medium 4020. When the computer-readable instructions 4010 are run by a processor, the method according to the embodiments of the present disclosure described with reference to the above figures can be executed. The computer-readable storage medium in the embodiments of the present disclosure can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. The non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory (DR RAM). It should be noted that the memory for the methods described herein is intended to include, but is not limited to, these and any other suitable types of memory. It should be noted that the memory for the methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0077] Embodiments of the present disclosure also provide a computer program product or a computer program. When the computer program is executed by a processor, the steps of the above method are implemented, as Figure 2 shown. The computer program product or the computer program includes: An acquisition module 201: configured to acquire an OCTA image of the eye of a diabetic patient; A feature extraction module 202: configured to extract choroidal features and retinal features based on the OCTA image, and calculate a choroid-to-retina feature ratio based on the ratio of the choroidal features and the retinal features; A prediction module 203: configured to input the choroid-to-retina feature ratio into a classifier to obtain a result of whether a diabetic patient has retinopathy.

[0078] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a portion of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system for performing the specified functions or operations, or may be implemented by a combination of dedicated hardware and computer instructions.

[0079] In general, the various example embodiments of the present disclosure may be implemented in hardware or a dedicated circuit, software, firmware, logic, or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software that can be executed by a controller, a microprocessor, or other computing devices. When aspects of the embodiments of the present disclosure are illustrated or described as block diagrams, flowcharts, or using some other graphical representation, it will be understood that the blocks, devices, systems, techniques, or methods described herein may be implemented as non-limiting examples in hardware, software, firmware, a dedicated circuit or logic, general hardware or a controller or other computing devices, or some combination thereof.

[0080] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above may refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0081] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the couplings or direct couplings or communication connections shown or discussed with each other may be indirect couplings or communication connections through some interfaces, devices, or units, and may be in electrical, mechanical, or other forms.

[0082] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0083] In addition, in each embodiment of the present invention, each functional unit may be integrated into a processing unit, may exist separately as individual physical units, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0084] The exemplary embodiments of the present disclosure described in detail above are merely illustrative and not restrictive. Those skilled in the art should understand that various modifications and combinations can be made to these embodiments or their features without departing from the principles and spirit of the present disclosure, and such modifications should fall within the scope of the present disclosure.

Claims

1. A prediction method for diabetic retinopathy based on the ratio of choroid-retina features, characterized in that, The method includes: Obtaining the ocular OCTA image of a diabetic patient; Extracting choroidal features and retinal features based on the ocular OCTA image, and calculating the choroid-to-retina feature ratio based on the ratio of the choroidal features and the retinal features; Inputting the choroid-to-retina feature ratio into a classifier to obtain the result of whether the diabetic patient has retinopathy.

2. The prediction method of diabetic retinopathy based on the choroid-retina feature ratio according to claim 1, wherein The choroid-to-retina feature ratio includes any one or more of the following: The ratio of the total choroidal volume to the total retinal volume, the ratio of the choroidal vascular volume to the total retinal volume, The ratio of the total choroidal volume to the outer retinal volume, the ratio of the choroidal vascular volume to the outer retinal volume.

3. The prediction method of diabetic retinopathy based on the choroid-retina feature ratio according to claim 2, wherein Extracting the choroidal features and retinal features of the central region where the fovea is located based on the ocular OCTA image, and calculating the choroid-to-retina feature ratio based on the ratio of the choroidal features and the retinal features of the central region.

4. The prediction method of diabetic retinopathy based on the choroid-retina feature ratio according to claim 3, characterized in that, The choroid-to-retina feature ratio includes any one or more of the following: the ratio of the total choroidal volume of the central region to the total retinal volume, the ratio of the choroidal vascular volume of the central region to the total retinal volume, the ratio of the total choroidal volume of the central region to the outer retinal volume, the ratio of the choroidal vascular volume of the central region to the outer retinal volume; Optionally, the choroid-to-retina feature ratio includes: the ratio of the total choroidal volume of the central region to the total retinal volume, where the total choroidal volume is the volume between the RPE / BM of the central region and the choroid-sclera interface, and the total retinal volume is the volume between the inner limiting membrane of the central region and the RPE / BM; Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroidal vascular volume of the central region to the total retinal volume, where the choroidal vascular volume is the volume of the large and medium-sized blood vessels in the choroid of the central region, and the total retinal volume is the volume between the inner limiting membrane of the central region and the RPE / BM; Optionally, the choroid-to-retina feature ratio includes: the ratio of the total choroidal volume of the central region to the outer retinal volume, where the total choroidal volume is the volume between the RPE / BM of the central region and the choroid-sclera interface, and the outer retinal volume is the volume between the outer plexiform layer of the central region and the RPE / BM; Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroidal vascular volume of the central region to the outer retinal volume, where the choroidal vascular volume is the volume of the large and medium-sized blood vessels in the choroid of the central region, and the outer retinal volume is the volume between the outer plexiform layer of the central region and the RPE / BM; Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroidal thickness to the retinal thickness of the central region, where the choroidal thickness is the average thickness between the RPE / BM of the central region and the choroid-sclera interface, and the retinal thickness is the average thickness between the inner limiting membrane of the central region and the RPE / BM; Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroid thickness in the central region to the outer retina thickness, where the choroid thickness is the average thickness between the RPE / BM in the central region and the choroid-scleral interface, and the outer retina volume is the average thickness between the outer plexiform layer and the RPE / BM in the central region.

5. The prediction method of diabetic retinopathy based on the choroid-retina feature ratio according to claim 2, wherein The ocular OCTA image is divided into a central region where the fovea is located and a peripheral region outside the fovea. The choroid features, retina features in the central region, and choroid features, retina features in the peripheral region are extracted respectively. The choroid-to-retina feature ratio is calculated based on the ratio of the choroid features and retina features in the central region, and the choroid-to-retina feature ratio in the peripheral region is calculated based on the ratio of the choroid features and retina features in the peripheral region. The choroid-to-retina feature ratio or the choroid-to-retina feature ratio in the peripheral region is input into the classifier to obtain the result of whether a diabetic patient has retinopathy. The choroid-to-retina feature ratio in the peripheral region includes any one or more of the following: the ratio of the total choroid volume in the peripheral region to the total retina volume, the ratio of the choroid vascular volume in the peripheral region to the total retina volume, the ratio of the total choroid volume in the peripheral region to the outer retina volume, the ratio of the choroid vascular volume in the peripheral region to the outer retina volume. Optionally, the choroid-to-retina feature ratio includes: the ratio of the total choroid volume in the peripheral region to the total retina volume; where the total choroid volume is the volume between the RPE / BM in the peripheral region and the choroid-scleral interface, and the total retina volume is the volume between the inner limiting membrane in the peripheral region and the RPE / BM. Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroid vascular volume in the peripheral region to the total retina volume, where the choroid vascular volume is the volume of large and medium-sized blood vessels in the choroid in the peripheral region, and the total retina volume is the volume between the inner limiting membrane in the peripheral region and the RPE / BM. Optionally, the choroid-to-retina feature ratio includes: the ratio of the total choroid volume in the peripheral region to the outer retina volume, where the total choroid volume is the volume between the RPE / BM in the peripheral region and the choroid-scleral interface, and the outer retina volume is the volume between the outer plexiform layer and the RPE / BM in the peripheral region. Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroid vascular volume in the peripheral region to the outer retina volume; where the choroid vascular volume is the volume of large and medium-sized blood vessels in the choroid in the peripheral region, and the outer retina volume is the volume between the outer plexiform layer and the RPE / BM in the peripheral region. Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroid thickness in the peripheral region to the retina thickness, where the choroid thickness is the average thickness between the RPE / BM in the peripheral region and the choroid-scleral interface, and the retina thickness is the average thickness between the inner limiting membrane in the peripheral region and the RPE / BM. Optionally, the choroid-to-retina feature ratio includes: the ratio of the choroid thickness in the peripheral region to the outer retinal thickness, where the choroid thickness is the average thickness between the RPE / BM in the peripheral region and the choroid-scleral interface, and the outer retinal volume is the average thickness between the outer plexiform layer and the RPE / BM in the peripheral region.

6. The prediction method of diabetic retinopathy based on the choroid-retina feature ratio according to claim 5, characterized in that Performing OCTA imaging in a scanning mode centered on the fovea to obtain the ocular OCTA image, and dividing the image into a 3×3 grid, where the central grid is designated as the central region and the surrounding grids are regarded as the peripheral regions; Optionally, concentric circles with different radii are set with the fovea center of the OCTA image as the center. The region within a radius of 6 mm is the central region, and the region between 6 mm and 12 mm is the peripheral region.

7. The prediction method of diabetic retinopathy based on the choroid-retina feature ratio according to claim 1, wherein The method for constructing the classifier is as follows: Obtaining the ocular OCTA images of diabetic patients in the training set and the corresponding labels, where the labels include no retinopathy and retinopathy; Extracting the choroid features and retinal features from the training set of ocular OCTA images, and calculating the choroid-to-retina feature ratio; Inputting the choroid-to-retina feature ratio and the labels into a machine learning model for iterative training to obtain the classifier.

8. A computer device, characterized in that, The device includes: a memory and a processor; the memory is used to store a computer program; the processor executes the computer program to implement the steps of the method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, A computer program is stored thereon, and when the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1-7.

10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1-7.

Images