深度学习去噪在肥胖患者低剂量冠状动脉CTA中的增量价值
本研究在56例（62.5%肥胖）低辐射冠状动脉CTA患者中，评估基于CNN的后处理去噪对图像质量的增益。去噪后图像噪声降低、SNR和CNR提升、主观质量改善，且去噪效率不依赖于BMI。
要点
- CNN去噪可显著降低低剂量CCTA图像噪声并提升SNR/CNR，且不改变血管衰减。
- 去噪后主观图像质量改善，差质量血管段比例从14.7%降至9.0%。
- 去噪效率与BMI无关，在肥胖患者中同样有效。
关键影像表现/征象
- CNN去噪降低图像噪声（35.1±7.8 vs 21.0±4.7 HU，p<0.001）。
- 去噪后SNR（12.0±3.3 vs 22.0±5.7，p<0.001）和CNR（11.0±3.0 vs 19.0±5.2，p<0.001）提升，衰减稳定（所有p>0.1）。
- 主观质量评分改善（2 [2,3] vs 3 [2,3]，p<0.001），差质量血管段从23/156（14.7%）降至14/156（9.0%），优良质量从82/156（52.6%）增至117/156（75.0%）。
- 混合效应模型显示噪声降低31%（-13.9 HU，p<0.001），SNR增益+8.5（p<0.001），CNR增益+7.4（p<0.001），与BMI无关（p>0.46）。
诊断标准与关键数值
- 主观图像质量采用4分法评估冠状动脉节段。
- 剂量长度乘积为77.15 [50.10-133.38] mGy·cm。
临床应用与阅片要点
- 对于低剂量CCTA图像，尤其肥胖患者，可考虑应用CNN去噪以改善图像质量而不影响血管衰减。
- 去噪效果不受BMI影响，可在肥胖患者中常规使用以提升诊断信心。
局限性
- 样本量较小（56例），且为单中心后处理研究。
- 仅评估了图像质量，未报告诊断准确性或临床结局。
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
OBJECTIVE: To determine the added value of post-hoc convolutional neural network (CNN)-based denoising for low-radiation coronary CT angiography (CCTA).
METHODS: In this IRB-approved, post-hoc study, consecutive patients underwent clinically indicated CCTA on a third-generation dual-source CT scanner using a fast-pitch protocol and fully patient-tailored contrast injection. Images were reconstructed with a vascular kernel (Bv40) using iterative reconstruction (ADMIRE 4), with and without CNN-based denoising. Two readers assessed signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), attenuation, and subjective image quality across coronary segments using a 4-point scale. Linear mixed-effects models evaluated the relationship between body-mass index (BMI) and denoising efficiency.
RESULTS: Fifty-six participants (mean age 51.7 ± 12.8 years; 57.1% women), including 35 obese individuals (62.5%, BMI ≥ 30.0 kg/m2), were analyzed. The dose-length product was 77.15 [50.10-133.38] mGy·cm. CNN-denoising reduced image noise (35.1 ± 7.8 vs 21.0 ± 4.7 HU, p < 0.001) and improved SNR (12.0 ± 3.3 vs 22.0 ± 5.7, p < 0.001) and CNR (11.0 ± 3.0 vs 19.0 ± 5.2, p < 0.001), with stable attenuation (all p > 0.1). Subjective quality improved (2 [2,3] vs 3 [2,3], p < 0.001). Poor-quality vessels decreased from 23/156 (14.7%) to 14/156 (9.0%), and good-to-excellent quality increased from 82/156 (52.6%) to 117/156 (75.0%). Mixed-effects modeling showed 31% noise reduction (-13.9 HU, p < 0.001) and SNR (+8.5, p < 0.001) and CNR (+7.4, p < 0.001) gains, irrespective of BMI (p > 0.46).
CONCLUSION: CNN-based denoising improved objective and subjective image quality in low-dose CCTA independently of BMI in a predominantly obese cohort.
引用
- Osoria-Velasquez J, Byerly K, Vecsey-Nagy M, et al. Incremental value of deep learning denoising in low-dose coronary CT angiography in predominantly obese patients. Eur J Radiol. 2026. PMID: 42001841. doi:10.1016/j.ejrad.2026.112859
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。