Abstract Background Diagnosis and follow-up of retinal diseases may be improved if the thickness of the various retinal layers, in addition to the total retinal thickness, is taken into account. Here we measured the thickness of the macular retinal layers in a population-based study group to assess the normative values and their associations. Methods Using spectral-domain optical coherence tomographic images, we measured the thickness of the macular retinal layers in participants of the population-based Beijing Eye Study without ocular diseases and without arterial hypertension, hyperlipidemia and diabetes mellitus. Results The study included 384 subjects (mean age:60.0±8.0 years). In multivariable analysis, the thickness of the retinal layers in the foveal region, of all retinal layers except for the outer plexiform layer in the parafoveal area, and the thickness of the ganglion cell layer, inner plexiform layer and inner and outer nuclear layer in the perifoveal area decreased with older age (all P<0.05). Men as compared to women had higher thickness measurements of the photoreceptor layer and outer nuclear layer in all areas, and of all layers between the retinal nerve fiber layer and inner nuclear layer in the parafovea area. The associations between the macular retinal layers thickness and axial length were not consistent. The inner plexiform layer was thicker, and the ganglion cell layer and inner nuclear layer were thinner, in the temporal areas than in the nasal areas, Conclusions The associations between decreasing thickness of most retinal layers with older age and the correlation of a higher thickness of some retinal layer layers with male gender may clinically be taken into account.
This study aimed to develop an automated computer-based algorithm to estimate axial length and subfoveal choroidal thickness (SFCT) based on color fundus photographs. In the population-based Beijing Eye Study 2011, we took fundus photographs and measured SFCT by optical coherence tomography (OCT) and axial length by optical low-coherence reflectometry. Using 6394 color fundus images taken from 3468 participants, we trained and evaluated a deep-learning-based algorithm for estimation of axial length and SFCT. The algorithm had a mean absolute error (MAE) for estimating axial length and SFCT of 0.56 mm [95% confidence interval (CI): 0.53,0.61] and 49.20 μm (95% CI: 45.83,52.54), respectively. Estimated values and measured data showed coefficients of determination of r 2 = 0.59 (95% CI: 0.50,0.65) for axial length and r 2 = 0.62 (95% CI: 0.57,0.67) for SFCT. Bland–Altman plots revealed a mean difference in axial length and SFCT of −0.16 mm (95% CI: −1.60,1.27 mm) and of −4.40 μm (95% CI, −131.8,122.9 μm), respectively. For the estimation of axial length, heat map analysis showed that signals predominantly from overall of the macular region, the foveal region, and the extrafoveal region were used in the eyes with an axial length of < 22 mm, 22–26 mm, and > 26 mm, respectively. For the estimation of SFCT, the convolutional neural network (CNN) used mostly the central part of the macular region, the fovea or perifovea, independently of the SFCT. Our study shows that deep-learning-based algorithms may be helpful in estimating axial length and SFCT based on conventional color fundus images. They may be a further step in the semiautomatic assessment of the eye.
Abstract Background Diabetic retinopathy and chronic kidney disease are both major complications of diabetes mellitus. We explored the relationship between retinal vessel density (VD) and albuminuria in diabetic patients without conventionally defined diabetic retinopathy. Methods The cross‐sectional community‐based Kailuan Diabetic Retinopathy Study included patients with type 2 diabetes without diabetic retinopathy who participated in the community‐based longitudinal Kailuan study and who had undergone ocular fundus photography, kidney function assessment, and optical coherence tomographic angiography (OCT‐angiography) for the assessment the retinal perfusion density (PD) and retinal VD. Results The study included 447 patients (mean age: 60.9 ± 9.7 years). Higher PD and VD were associated with a lower urinary albumin‐creatinine ratio (uACR) (macular region: p = 0.007: standardized regression coefficient beta: −0.14; and p = 0.008, beta: −0.13, respectively; parafoveal region: p = 0.006, beta: −0.14; and p = 0.007, beta: −0.14, respectively) after adjusting for age and ocular axial length. In a reverse manner, higher uACR was associated with lower PD and VD (macular region: p = 0.009, beta: −0.14; and p = 0.01, beta: −0.14, respectively; parafoveal region: p = 0.008, beta: −0.14; and p = 0.01, beta: −0.14, respectively), after adjusting for diabetes duration, blood pressure, serum concentration of C‐reactive protein and high‐density lipoprotein cholesterol and ocular axial length. In a multivariable model, the prevalence of macroalbuminuria increased by 11% (95% CI: 2%, 18%) and 17% (95% CI: 3%, 30%), respectively, for each mm −1 decrease in VD and each unit decrease in PD. Conclusions and Relevance After adjusting for systemic and ocular parameters, diabetic patients without diabetic retinopathy showed a reduction in OCT‐angiographic retinal vascular measurements in association with systemic parameters indicating chronic kidney disease. Optical coherence tomographic (OCT)‐angiographic retinal microvascular parameters may serve as markers for chronic kidney disease.
Purpose: To examine the role of ocular axial length as an ocular parameter for the prevalence and severity of diabetic retinopathy (DR). Methods: The cross-sectional Kailuan Diabetic Retinopathy Study included patients with diabetes who participated in the community-based longitudinal Kailuan Study and who had undergone ocular fundus photography. The fundus photographs were graded using the Early Treatment of Diabetic Retinopathy Study criteria. Results: The study included 1096 patients with diabetes (mean age: 60.8 ± 9.4 years; axial length: 23.37 ± 0.92 mm). In binary regression analysis, a higher DR prevalence was associated with shorter axial length (P = 0.007; odds ratio [OR]: 0.81; 95% confidence interval [CI]: 0.70, 0.95) after adjusting for longer known duration of diabetes (P = 0.02; OR: 1.13; 95%CI: 1.02, 1.24) and higher fasting blood glucose concentration (P < 0.001; OR: 1.38; 95%CI: 1.26, 1.52). A more severe DR stage was associated (regression coefficient r: 0.46) with shorter ocular axial length (P = 0.047; standardized regression coefficient β: −0.06) after adjusting for higher fasting blood glucose (P < 0.001; β: 0.41) and longer known duration of diabetes (P = 0.045; β: 0.07). Longer axial length was associated with a lower DR prevalence (P = 0.003; β: −0.10) after adjusting for younger age (P < 0.001), male sex (P < 0.001), higher body mass index (P = 0.016), and lower fasting blood glucose concentration (P = 0.036). Conclusions: After adjusting for systemic risk factors, DR prevalence decreased by 19% (95%CI: 5, 30) for each millimeter increase in axial length. With longer axial length being a surrogate for axial myopia, the marked increase in myopia prevalence worldwide may lead to a relative decrease in the prevalence and incidence of DR in future.
To assess the progression of fundus tessellation, color fundus photographs of the participants of the longitudinal population-based Beijing Eye Study were examined. The study included 4439 subjects in 2001 and 2695 (66.4% of the surviving) individuals in 2011. Larger progression in macular fundus tessellation (mean: 0.24 ± 0.48 grades) was associated (multivariate analysis; correlation coefficient r: 0.53) with thinner subfoveal choroidal thickness in 2011 (P < 0.001; standardized regression coefficient beta: -0.37), older age (P < 0.001; beta: 0.22), higher level of education (P < 0.001; beta: 0.09), more myopic change in refractive error (P < 0.001; beta: -0.09) and lower cognitive function score (P = 0.02; beta: -0.05). Larger increase in peripapillary fundus tessellation (mean: 0.19 ± 0.26 grades) correlated with thinner peripapillary choroidal thickness in 2011 (P < 0.001; beta: -0.35), older age (P < 0.001; beta: 0.20), worse best corrected visual acuity (P = 0.001; beta: 0.07), more myopic change in refractive error (P < 0.001; beta: -0.07) and higher prevalence of ever smoking (P = 0.004; beta: 0.05). The increase in macular fundus tessellation, as a surrogate for thinning of the posterior choroid, was associated with lower cognitive function, after adjusting for choroidal thickness, age, educational level and change in refractive error. The findings point to the clinical value of the assessment of fundus tessellation and suggest potential associations between cognitive function and fundus tessellation/choroidal thickness.
Abstract The aim of the study was to assess longitudinal changes in the spatial relationship of the choroidal vasculature to retinal vasculature in myopic eyes. In the population-based longitudinal Beijing Eye Study in 2001/2011, we examined all highly myopic eyes with assessable fundus photographs and a randomized group of non-highly myopic. Using fundus photographs, we qualitatively assessed changes in the location of major choroidal vessels in relationship to retinal vessels. The study consisted of 85 highly myopic eyes (58 participants;age:64.8 ± 9.4 years) and 85 randomly selected non-highly myopic eyes. A choroidal shift in relationship to the retinal vessels was detected more often in the highly myopic group than the non-highly myopic group (47/85 (55%) vs 6/85 (7%); P < 0.001). In the highly myopic group, the choroidal vessel shift occurring on the disc-fovea line in 39 (44%) eyes, was similar to, or smaller than, the enlargement in gamma zone width in 26 (67%) eyes and in 11 (28%) eyes respectively. The choroidal vessel shift was larger ( P = 0.002) in eyes without choroidal vessels in gamma zone than in eyes with large choroidal vessels in gamma zone. In 14 (17%) eyes, a localized centrifugal choroidal shift was observed in association with an increase in the stage of myopic maculopathy. The results suggest that highly myopic eyes show a change in the position of large choroidal vessels in relationship to retinal vessels, in association with development or enlargement of gamma zone and an increase in the stage of myopic maculopathy.
We aimed to develop a deep learning system capable of identifying subjects with cognitive impairment quickly and easily based on multimodal ocular images. Cross-sectional study Participants of Beijing Eye Study 2011 and patients attending Beijing Tongren Eye Center and Beijing Tongren Hospital Physical Examination Center. We trained and validated a deep learning algorithm to assess cognitive impairment using retrospectively collected data from the Beijing Eye Study 2011. Cognitive impairment was defined as a Mini–Mental State Examination (MMSE) score <24. Based on fundus photographs and optical coherence tomography (OCT) images, we developed five models based on the following sets of images: macula-centered fundus photographs, optic disc-centered fundus photographs, fundus photographs of both fields, optical coherence tomography (OCT) images, and fundus photographs of both fields with OCT (multi-modal). The performance of the models was evaluated and compared in an external validation dataset, which was collected from patients attending Beijing Tongren Eye Center and Beijing Tongren Hospital Physical Examination Center. Area under the curve (AUC). A total of 9,424 retinal photographs and 4,712 OCT images were used to develop the model. The external validation sets from each center included 1,180 fundus photographs and 590 OCT images. Model comparison revealed that the multi-modal performed best, achieving an AUC of 0.820 in the internal validation set, 0.786 in external validation set 1 and 0.784 in external validation set 2. We evaluated the performance of the multi-model in different sexes and different age groups; there were no significant differences. The heatmap analysis showed that signals around the optic disc in fundus photographs and the retina and choroid around the macular and optic disc regions in OCT images were used by the multi-modal to identify participants with cognitive impairment. Fundus photographs and OCT can provide valuable information on cognitive function. Multi-modal models provide richer information compared to single-mode models. Deep learning algorithms based on multimodal retinal images may be capable of screening cognitive impairment. This technique has potential value for broader implementation in community-based screening or clinic settings.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
