ObjectiveTo build a small-sample ultra-widefield fundus images (UWFI) multi-disease classification artificial intelligence model, and initially explore the ability of artificial intelligence to classify UWFI multi-disease tasks. MethodsA retrospective study. From 2016 to 2021, 1 608 images from 1 123 patients who attended the Eye Center of the Renmin Hospital of Wuhan University and underwent UWFI examination were used for UWFI multi-disease classification artificial intelligence model construction. Among them, 320, 330, 319, 268, and 371 images were used for diabetic retinopathy (DR), retinal vein occlusion (RVO), pathological myopia (PM), retinal detachment (RD), and normal fundus images, respectively. 135 images from 106 patients at the Tianjin Medical University Eye Hospital were used as the external test set. EfficientNet-B7 was selected as the backbone network for classification analysis of the included UWFI images. The performance of the UWFI multi-task classification model was assessed using the receiver operating characteristic curve, area under the curve (AUC), sensitivity, specificity, and accuracy. All data were expressed using numerical values and 95% confidence intervals (CI). The datasets were trained on the network models ResNet50 and ResNet101 and tested on an external test set to compare and observe the performance of EfficientNet with the 2 models mentioned above. ResultsThe overall classification accuracy of the UWFI multi-disease classification artificial intelligence model on the internal and external test sets was 92.57% (95%CI 91.13%-92.92%) and 88.89% (95%CI 88.11%-90.02%), respectively. These were 96.62% and 92.59% for normal fundus, 95.95% and 95.56% for DR, 96.62% and 98.52% for RVO, 98.65% and 97.04% for PM, and 97.30% and 94.07% for RD, respectively. The mean AUC on the internal and external test sets was 0.993 and 0.983, respectively, with 0.994 and 0.939 for normal fundus, 0.999 and 0.995 for DR, 0.985 and 1.000 for RVO, 0.991 and 0.993 for PM and 0.995 and 0.990 for RD, respectively. EfficientNet performed better than the ResNet50 and ResNet101 models on both the internal and external test sets. ConclusionThe preliminary UWFI multi-disease classification artificial intelligence model using small samples constructed in this study is able to achieve a high accuracy rate, and the model may have some value in assisting clinical screening and diagnosis.
ObjectiveTo observe clinical phenotypes and analyze the pathogenic genes of Leber congenital amaurosis (LCA). MethodsA retrospective clinical study. From 2019 to 2020, 2 patients diagnosed with LCA by genetic testing in Tianjin Medical University Eye Hospital and their 6 unaffected family members were enrolled in the study. Two patients were from 2 unrelated families, both were probands. The patient's medical history was inquired in detail, slit lamp microscopy, ultra-widefield fundus photography, autofluorescence, and flash visual evoked potential (F-VEP) were performed. Peripheral vein blood (3-5 ml) was collected and genomic DNA was extracted from all study subjects. A total of 381 pathogetic genes associated with inherited retinal diseases, were selected by targeted exome sequencing capture strategy. Sanger sequencing was used to verify suspected pathogenic mutations. Candidate pathogenic mutations were identified after bioinformatics analysis. Sanger sequencing, real-time quantitative polymerase chain reaction and family co-identification were used to confirm the final mutations. ResultsTwo patients were male, aged 3 and 27 years. One case had vision loss in both eyes, accompanied by nystagmus and acupressure eye sign since childhood. The clinical hallmark of the proband (F1-Ⅱ-3) in F1 includes clearly boundary of optic disc, normal retinal blood vessels and macular fovea. The implied period of the maximum forward wave in both eyes of F-VEP was roughly normal, and its amplitude decreased significantly. The phenotype of the proband (F2-Ⅱ-1) in F2 includes optic nerve head pallor, bone-spicule intraretinal pigmentation, “gold-foil maculopathy”, retina patchy hypo-autofluorescence in both eyes. There was no abnormal phenotype in the eyes of the family members. According to the genetic diagnosis, the proband (F1-Ⅱ-3) carried the GUCY2D gene c.835G>A (p.D279N) (M1) and exon 9-19 deletion (M2) compound heterozygous mutations, in which M1 was derived from healthy mother and M2 was derived from healthy father. The proband (F2-Ⅱ-1) carried CRB1 gene c.1576C>T(R526X) (M3) and c.1522T>C (C508R) (M4) compound heterozygous mutations, in which M3 from the healthy father, M4 from the healthy mother. M2 and M4 were novel mutations. ConclusionGUCY2D gene mutations lead to LCA1 type in the F1 family, CRB1 gene mutations lead to LCA8 type in the F2 family; there are significant different phenotypes caused by different pathogenic genes.
ObjectiveTo observe the effect of interleukin-8 (IL-8) on the adhesion and migration of retinal vascular endothelial cells (RCEC). MethodsA cell experiment. Human RCEC (hRCEC) was divided into normal control group (N group), advanced glycation end product (AGE) treatment group (AGE group), and AGE-induced combined IL-8 antagonist SB225002 treatment group (AGE+SB group). The effect of AGE on IL-8 expression in hRCEC was observed by Western blot. The effect of SB225002 on hRCEC migration was observed by cell scratch assay. The effects of SB225002 on leukocyte adhesion and reactive oxygen species (ROS) on hRCEC were detected by flow cytometry. Student-t test was performed between the two groups. One-way analysis of variance was performed among the three groups. ResultsCompared with group N, the expression level of IL-8 in cells of AGE group was significantly increased, with statistical significance (t=25.661, P＜0.001). Compared with N group and AGE+SB group, cell mobility in AGE group was significantly increased (F=29.776), leukocyte adhesion number was significantly increased (F=38.159, 38.556), ROS expression level was significantly increased (F=22.336), and the differences were statistically significant (P＜0.05). ConclusionIL-8 antagonist SB225002 may down-regulate hRCEC adhesion and migration by inhibiting ROS expression.