Incidence rates of retinal vascular occlusive diseases from 2011 to 2020 in South Korea: a nationwide cohort study.

BACKGROUND: Retinal vascular occlusions, including retinal vein occlusion and retinal artery occlusion, are common causes of visual impairment. In order to evaluate the national medical burden and help improve ophthalmic health care policy planning, we investigated the incidence of retinal vascular occlusive diseases from 2011 to 2020 in Korea. METHODS: This study is a nationwide population-based retrospective study using data from the Korea national health claim database of the Health Insurance Review and Assessment (HIRA) service. We identified retinal vascular occlusive diseases registered from January 1, 2009, to December 31, 2020, according to the retinal vascular occlusion code (H34) and its sub-codes from international classification of disease, tenth revision diagnosis code. We used data from the entire Korean population based on the 2015 census of the population in Korea to calculate standardized incidence rates. RESULTS: We identified 348,775 individuals (male, 161,673 [46.4%]; female, 187,102 [53.6%]) with incident retinal vascular occlusion (H34), 10,451 individuals (males, 6,329 [60.6%]; females, 4,122 [39.4%]) with incident central retinal artery occlusion (H34.1), and 252,810 individuals (males, 114,717 [45.4%]; females, 138,093 [54.6%]) with incident retinal vein occlusion (H34.8) during the 10-year study period. The weighted mean incidence rate of retinal vascular occlusion was 70.41 (95% CI, 70.18-70.65) cases/100,000 person-years. The weighted mean incidence rate of central retinal artery occlusion was 2.10 (95% CI, 2.06-2.14) cases/100,000 person-years. The weighted mean incidence rate of retinal vein occlusion was 50.99 (95% CI, 50.79-51.19) cases/100,000 person-years. CONCLUSION: The total retinal vascular occlusion and retinal vein occlusion showed a decreasing trend until 2020. However, the central retinal artery occlusion decreased until 2014 and remained stable without a significant further decline until 2020. The incidence of total retinal vascular occlusion and retinal vein occlusion was higher in females than in males, while the incidence of central retinal artery occlusion was higher in males. All retinal vascular occlusive diseases showed an increasing incidence with older age; the peak age incidence was 75-79 years for total retinal vascular occlusion and retinal vein occlusion, and 80-85 years for central retinal artery occlusion.

Asymmetry between right and left optical coherence tomography images identified using convolutional neural networks.

In a previous study, we identified biocular asymmetries in fundus photographs, and macula was discriminative area to distinguish left and right fundus images with > 99.9% accuracy. The purposes of this study were to investigate whether optical coherence tomography (OCT) images of the left and right eyes could be discriminated by convolutional neural networks (CNNs) and to support the previous result. We used a total of 129,546 OCT images. CNNs identified right and left horizontal images with high accuracy (99.50%). Even after flipping the left images, all of the CNNs were capable of discriminating them (DenseNet121: 90.33%, ResNet50: 88.20%, VGG19: 92.68%). The classification accuracy results were similar for the right and left flipped images (90.24% vs. 90.33%, respectively; p = 0.756). The CNNs also differentiated right and left vertical images (86.57%). In all cases, the discriminatory ability of the CNNs yielded a significant p value (< 0.001). However, the CNNs could not well-discriminate right horizontal images (50.82%, p = 0.548). There was a significant difference in identification accuracy between right and left horizontal and vertical OCT images and between flipped and non-flipped images. As this could result in bias in machine learning, care should be taken when flipping images.