ABSTRACT
Objective:To apply the multi-modal deep learning model to automatically classify the ultra-widefield fluorescein angiography (UWFA) images of diabetic retinopathy (DR).Methods:A retrospective study. From 2015 to 2020, 798 images of 297 DR patients with 399 eyes who were admitted to Eye Center of Renmin Hospital of Wuhan University and were examined by UWFA were used as the training set and test set of the model. Among them, 119, 171, and 109 eyes had no retinopathy, non-proliferative DR (NPDR), and proliferative DR (PDR), respectively. Localization and assessment of fluorescein leakage and non-perfusion regions in early and late orthotopic images of UWFA in DR-affected eyes by jointly optimizing CycleGAN and a convolutional neural network (CNN) classifier, an image-level supervised deep learning model. The abnormal images with lesions were converted into normal images with lesions removed using the improved CycleGAN, and the difference images containing the lesion areas were obtained; the difference images were classified by the CNN classifier to obtain the prediction results. A five-fold cross-test was used to evaluate the classification accuracy of the model. Quantitative analysis of the marker area displayed by the differential images was performed to observe the correlation between the ischemia index and leakage index and the severity of DR.Results:The generated fake normal image basically removed all the lesion areas while retaining the normal vascular structure; the difference images intuitively revealed the distribution of biomarkers; the heat icon showed the leakage area, and the location was basically the same as the lesion area in the original image. The results of the five-fold cross-check showed that the average classification accuracy of the model was 0.983. Further quantitative analysis of the marker area showed that the ischemia index and leakage index were significantly positively correlated with the severity of DR ( β=6.088, 10.850; P<0.001). Conclusion:The constructed multimodal joint optimization model can accurately classify NPDR and PDR and precisely locate potential biomarkers.
ABSTRACT
Objective:To analyze the changes of subfoveal choroidal thickness (SFCT) and choroidal volume in type 2 diabetes patients with diabetic retinopathy (DR) through automated choroid segmentation.Methods:A cross-sectional study was conducted.Sixty-four patients (88 eyes) with type 2 diabetes diagnosed in The First Affiliated Hospital with Nanjing Medical University from May 2016 to May 2017 were enrolled, including 27 males (32 eyes) and 37 females (56 eyes), with an average age of (62.7±7.4) years and an average diabetes duration of (13.7±7.2) years.Best corrected visual acuity, slit lamp ophthalmoscopy, indirect ophthalmoscopy and spectral domain optical coherence tomography (SD-OCT) examination were carried out among all the patients.The patients were divided into non-DR (NDR) group (28 eyes), nonproliferative DR (NPDR) group (36 eyes) and proliferative DR (PDR) group (24 eyes) according to the Early Treatment Diabetic Retinopathy Study.Nineteen age-matched normal subjects (32 eyes) were enrolled as normal control group.SFCT and choroidal volume were measured through automated choroid segmentation based on SD-OCT with enhanced depth imaging.This study adhered to the Declaration of Helsinki.The study protocol was approved by the Ethics Committee of The First Affiliated Hospital with Nanjing Medical University (No.2017-SR-176).Results:Mean SFCT and choroidal volume of normal control group, NDR group, NPDR group and PDR group were (277.8±45.3)μm and (8.8±1.0)mm 3, (220.9±38.7)μm and (7.6±1.3)mm 3, (231.8±26.7)μm and (8.2±0.9)mm 3, (252.0±42.2)μm and (8.7±1.0)mm 3, respectively, with significant overall differences among the groups ( F=14.615, P<0.001; F=7.711, P<0.001). Mean SFCT and choroidal volume were significantly smaller in NDR and NPDR group than those in normal control group (both at P<0.05). Mean SFCT and choroidal volume of PDR group was greater than those of NDR group (both at P<0.05). Conclusions:SFCT and choroidal volume decrease during the early course of diabetes and increase significantly as DR worsens from NDR to PDR.
ABSTRACT
Replacement of diseased retinal pigment epithelium (RPE) cells with healthy RPE cells by transplantation is one option to treat several retinal degenerative diseases including age-related macular degeneration,which are caused by RPE loss and dysfunction.A cellular scaffold as a carrier for transplanted cells,may hold immense promise for facilitating cell migration and promoting the integration of RPE ceils into the host environment.Scaffolds can be prepared from a variety of natural and synthetic materials.Strategies,such as surface modification and structure adjustment,can improve the biomimetic properties of the scaffolds,optimize cell attachment and cellular function following transplantation and lay a foundation of clinical application in the future.
ABSTRACT
Based on the pathogenic mechanisms of age-related macular degeneration (AMD),tremendous preclinical and clinical trials have demonstrated that cell transplantation which aim to replace impaired retinal pigment epithelium (RPE) with healthy RPE cells is a promising approach to treat AMD.So far,choices of cell sources mainly are autologous RPE,iris pigment epithelium,fetal RPE,human embryonic stem cell-derived RPE and human induced pluripotent stem cell-derived RPE,and some of them are undergoing clinical researches.Grafting manners in cell-based therapies are various including RPE sheet or RPE-choroid complex transplantation,RPE cell suspension injection,and RPE sheet transplantation with scaffolds.This review is limited to cell-based therapies for RPE that damaged first in the progress of AMD and focus on recent advances in cell sources,transplantation methods,preclinical and clinical trials,and the obstacles that must be overcome.