Key result
Fusion-YOLOv11-L improves coronary stenosis detection accuracy to an AP@50 of ~0.59 over baseline YOLOv11-L.
Why the study?
Manual interpretation of invasive X-ray coronary angiography for stenosis detection is time-consuming, subjective, and variable, while existing deep learning methods often rely solely on either detection or segmentation, limiting performance.
Does the Fusion-YOLOv11-L hybrid deep learning framework improve automated coronary artery stenosis detection (AP@50) in invasive X-ray coronary angiography images compared to baseline YOLOv11-L?
Does the Fusion-YOLOv11-L hybrid deep learning framework improve automated coronary artery stenosis detection (AP@50) in invasive X-ray coronary angiography images compared to baseline YOLOv11-L?
Fusing vessel-aware segmentation features from UNet++ into a YOLOv11-L detector via channel attention gating, combined with test-time augmentation, improves the accuracy and cross-domain generalisability of automated coronary stenosis detection in X-ray angiography.
Answers are grounded in this paper and cite their sources.
May enhance AI stenosis detection in angiography; hypothesis-generating for clinical adoption pending prospective validation.
Coronary artery disease remains a leading cause of mortality, with stenosis detection from invasive X-ray Coronary Angiography (XCA) critical for diagnosis. Manual interpretation is time-consuming, subjective, and variable, especially under low contrast and vessel overlap. Existing deep learning methods often rely solely on detection or segmentation, limiting performance. We propose a hybrid framework, Fusion-YOLOv11-L, combining a YOLOv11-L detector with a pretrained UNet++ via a channel attention gating module. This enables fusion of semantic and vessel-aware structural features, improving localisation of stenotic regions. The UNet++ is pretrained on ARCADE vessel segmentation dataset to provide anatomically informed representations. Experiments on the relabeled ARCADE (ARCADE*) dataset show improved performance over YOLOv11-L, achieving AP@50 of 0.585 and F1-score of 0.596. Cross-domain evaluation on a dataset from the Oxford Acute Myocardial Infarction (OxAMI) Study reveals performance degradation due to domain shift, mitigated using test-time adaptation and augmentation. Test-time augmentation improves AP@50 from 0.434 to 0.480, indicating better generalisability. These results show that cross-architecture feature fusion and lightweight adaptation improve robustness and generalisability of automated stenosis detection in XCA.
No takes yet. Share an insight, caveat, or question.
Zhu et al. (2026) studied Coronary artery disease / Coronary artery stenosis (n=1,527). Fusion-YOLOv11-L (hybrid deep learning framework combining YOLOv11-L with pretrained UNet++ via channel attention gating) vs. Baseline YOLOv11-L detector was evaluated on Average Precision at an Intersection-over-Union threshold of 0.5 (AP@50) for stenosis detection on the in-domain ARCADE test set. Fusion-YOLOv11-L improved coronary stenosis detection AP@50 to 0.585 (vs 0.574 for YOLOv11-L) on ARCADE, and test-time augmentation increased out-of-distribution AP@50 from 0.434 to 0.480.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: