Relationship between ischemic index, leakage index, and macular edema in branch retinal vein occlusion.

PURPOSE: To investigate the combined association of the ischemic index and leakage index with macular edema on ultra-widefield fluorescein angiography (UWFFA) in patients with branch retinal vein occlusion (BRVO). METHODS: Retrospective image analysis study. The leakage index and ischemic index were calculated using Fiji after aligning early and late UWFFA images. Differences in the ischemic index, leakage index, and central macular thickness (CMT) between ischemic and non-ischemic BRVO were compared. Moreover, the association between the ischemic index, leakage index, and macular edema was analyzed. RESULTS: Eighty-three patients with BRVO were enrolled, including 53 non-ischemic BRVO and 30 ischemic BRVO patients. No significant differences were observed in leakage index and CMT between ischemic BRVO and non-ischemic BRVO (all P > 0.05). In all included patients, CMT correlated with the panretina and all subregion leakage indexes (all P < 0.01), but not with the ischemic index (all P > 0.05). In the ischemic BRVO group, CMT showed a correlation with the leakage index in several regions, but not with the ischemic index. After adjusting for the ischemic index and other clinical features, CMT remained significantly correlated with the leakage index in all regions. CONCLUSION: The leakage index may be a more effective biomarker for monitoring BRVO-associated macular edema compared to the ischemic index. Further follow-up studies are warranted to validate these findings.

Automatic Detection of 30 Fundus Diseases Using Ultra-Widefield Fluorescein Angiography with Deep Experts Aggregation.

INTRODUCTION: Inaccurate, untimely diagnoses of fundus diseases leads to vision-threatening complications and even blindness. We built a deep learning platform (DLP) for automatic detection of 30 fundus diseases using ultra-widefield fluorescein angiography (UWFFA) with deep experts aggregation. METHODS: This retrospective and cross-sectional database study included a total of 61,609 UWFFA images dating from 2016 to 2021, involving more than 3364 subjects in multiple centers across China. All subjects were divided into 30 different groups. The state-of-the-art convolutional neural network architecture, ConvNeXt, was chosen as the backbone to train and test the receiver operating characteristic curve (ROC) of the proposed system on test data and external test date. We compared the classification performance of the proposed system with that of ophthalmologists, including two retinal specialists. RESULTS: We built a DLP to analyze UWFFA, which can detect up to 30 fundus diseases, with a frequency-weighted average area under the receiver operating characteristic curve (AUC) of 0.940 in the primary test dataset and 0.954 in the external multi-hospital test dataset. The tool shows comparable accuracy with retina specialists in diagnosis and evaluation. CONCLUSIONS: This is the first study on a large-scale UWFFA dataset for multi-retina disease classification. We believe that our UWFFA DLP advances the diagnosis by artificial intelligence (AI) in various retinal diseases and would contribute to labor-saving and precision medicine especially in remote areas.