视网膜血管标志物概念感知与多模态融合的冠状动脉病变检测方法
By constructing a cross-sectional dataset of coronary artery lesions and a conceptual awareness model of retinal vascular markers, we achieved multimodal fusion of retinal vascular morphology and clinical characteristics, solving the invasiveness problem of existing coronary artery lesion diagnosis, improving the accuracy and interpretability of detection, and providing a non-invasive diagnostic method.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NORTHEAST FORESTRY UNIV
- Filing Date
- 2025-09-28
- Publication Date
- 2026-07-17
AI Technical Summary
Current diagnosis of coronary artery disease mainly relies on highly invasive coronary angiography, and the use of existing retinal vascular markers is insufficient, resulting in inadequate accuracy and interpretability of test results.
A cross-sectional dataset of coronary artery lesions was constructed, retinal fundus images and clinical information were collected, vascular markers were extracted using a retinal arteriovenous segmentation model, vascular morphological changes were quantified through a biomarker concept perception module, and vascular features and clinical features were fused by a clinical feature representation module and a cross-modal attention aggregation module. The model was trained until convergence for the diagnosis of coronary artery lesions.
It improves the accuracy and interpretability of coronary artery disease diagnosis, provides a non-invasive diagnostic method, and reduces the burden on patients.
Smart Images
Figure CN121074010B_ABST