Cross-modality labeling enables non-invasive capillary quantification as a sensitive biomarker for assessing cardiovascular risk

Abstract

Purpose: We aim to use Fundus fluorescein angiography (FFA) to label the capillaries on color fundus photographs (CF) and train a deep learning model to quantify retinal capillaries non-invasively from CF and apply it to cardiovascular disease (CVD) risk assessment. Design: cross-sectional study Subjects: 90,732 pairs of CF-FFA images from 3893 participants. Main Outcome Measures: Area under the ROC curve (AUC), accuracy, sensitivity, and specificity for segmentation. Hazard ratio [95% confidence interval (CI)] for Cox regression analysis. Method: We matched the vessels extracted from FFA and CF, and used vessels from FFA as labels to train a deep learning model (RMHAS-FA) to segment retinal capillaries using CF. We tested the model's accuracy on a manually labeled internal test set (FundusCapi). For external validation, we tested the segmentation model on seven vessel segmentation datasets, and investigated the clinical value of the segmented vessels in predicting CVD events using data from 49,229 participants in the UK Biobank. Results: On the FundusCapi dataset, the segmentation performance was AUC = 0.94, accuracy = 0.93, sensitivity = 0.89, and specificity = 0.93. Smaller vessel skeleton density had a stronger correlation with CVD risk factors and incidence (p < 0.01). Reduced density of small vessel skeletons was strongly associated with an increased risk of CVD incidence and mortality for women (HR [95% CI] = 0.91[0.84-0.98] and 0.68[0.54-0.86], respectively). Conclusions: Using paired CF-FFA images, we automated the laborious manual labeling process and enabled non-invasive capillary quantification from CF, supporting its potential as a sensitive screening method for identifying individuals at high risk of future CVD events.