深度学习自动冠状动脉钙化评分:跨多种胸部CT协议的验证
本研究在一所大学医院回顾性收集2013至2023年2132例无动脉粥样硬化性心血管疾病患者的胸部CT(常规、CT-CAC及CT-COVID协议),训练并验证基于深度学习的冠状动脉钙化自动分割模型,以有经验观察者手动标注为参考标准。模型Agatston评分ICC达0.987,钙化风险分类Cohen kappa为0.86至0.95,CAC>100与CAC>300的F1分别为0.956与0.967,外部验证Mashhad COVID研究kappa为0.8,SBU COVID研究中重度钙化F1为0.928。
要点
- 深度学习模型自动量化冠状动脉钙化,Agatston评分与人工标注高度一致(ICC=0.987)。
- 钙化风险分类一致性Cohen kappa为0.86至0.95,CAC>100与CAC>300的F1分别为0.956与0.967。
- 模型跨常规、专用CT-CAC及COVID时期胸部CT协议均表现良好,外部验证kappa为0.8。
核心结论
- 所提出的深度学习模型在常规、专用及疫情时期胸部CT扫描中均能实现准确、可直接用于工作流程的冠状动脉钙化量化,支持机会性、成本效益好的心血管风险分层。
研究设计与样本
- 回顾性研究,在一所大学医院收集2013至2023年2132例无确立动脉粥样硬化性心血管疾病患者的胸部CT(常规、CT-CAC和CT-COVID协议),训练并验证深度学习冠状动脉钙化分割模型,以有经验观察者手动标注为参考标准。
诊断标准与关键数值
- Agatston评分与人工标注的组内相关系数ICC=0.987。
- 钙化风险分类一致性Cohen kappa=0.86至0.95。
- CAC>100的F1=0.956,CAC>300的F1=0.967。
- 外部验证Mashhad COVID研究kappa=0.8,SBU COVID研究中重度钙化检测F1=0.928。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
RATIONALE AND OBJECTIVES: Coronary artery calcium (CAC) scoring refines atherosclerotic cardiovascular disease (ASCVD) risk but is not frequently reported on routine non‑gated chest CT (NCCT), whose use expanded in the COVID‑19 era. We sought to develop and validate a workflow-ready deep learning model for fully automated, protocol-agnostic CAC quantification.
MATERIALS AND METHODS: In this retrospective study, a deep learning (DL) model was trained and validated using 2132 chest CT scans (routine, CT-CAC, and CT-COVID) from patients without established atherosclerotic cardiovascular disease (ASCVD) collected (2013-2023) at a single university hospital. The index test was a DL-based CAC segmentation model; the reference standard was manual annotation by experienced observers. Agreement was evaluated using intra-class correlation coefficients (ICC) for Agatston scores and Cohen's kappa for CAC risk categories. Sensitivity, specificity, positive, and negative predictive values, and F1 scores were calculated to measure diagnostic performance.
RESULTS: The DL model demonstrated high reliability for Agatston scores (ICC=0.987) and strong agreement in CAC categories (Cohen's κ=0.86-0.95). Diagnostic performance for CAC >100 (F1=0.956) and CAC >300 (F1=0.967) was very high. External validation in the Mashhad COVID Study showed good agreement (κ=0.8). In the SBU COVID study, the F1 score for detecting moderate-to-severe CAC was 0.928.
CONCLUSION: The proposed DL model delivers accurate, workflow‑ready CAC quantification across routine, dedicated, and pandemic‑era chest CT scans, supporting opportunistic, cost‑effective cardiovascular risk stratification in contemporary clinical practice.
引用
- Mineo E, Assuncao-Jr AN, Grego da Silva CF, et al. Automated Coronary Artery Calcium Scoring Using Deep Learning: Validation Across Diverse Chest CT Protocols. Acad Radiol. 2025. PMID: 40998657. doi:10.1016/j.acra.2025.09.009
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。