| 张卓璐,安备,商旭,刘卓,王屹,洪楠.重建算法对于冠状动脉CT量化分析的影响[J].中国医学影像技术,2026,42(3):444~447 |
| 重建算法对于冠状动脉CT量化分析的影响 |
| Impact of reconstruction algorithm on coronary artery CT quantitative analysis |
| 投稿时间：2025-07-16 修订日期：2025-11-06 |
| DOI：10.13929/j.issn.1003-3289.2026.03.025 |
| 中文关键词: 深度学习 冠状血管 体层摄影,X线计算机 斑块 |
| 英文关键词:deep learning coronary vessels tomography,X-ray computed plaque |
| 基金项目: |
|
| 摘要点击次数: 69 |
| 全文下载次数: 26 |
| 中文摘要: |
| 目的 观察重建算法对于冠状动脉CT量化分析的影响。方法 收集58例患者的原始冠状动脉CT(包括钙化积分扫描及血管成像)数据,采用自适应统计迭代重建(ASIR-V)和深度学习图像重建(DLIR),分别以不同权重ASIR-V(ASIR-V0%、ASIR-V50%、ASIR-V100%)及低、中、高级别DLIR(DLIR-L、DLIR-M、DLIR-H)重建图像,测量其中的钙化评分(Agatston积分、质量积分、体积积分)、斑块,以及管腔量化分析结果、血管周围脂肪CT值及升主动脉CT值并进行比较。结果 各重建算法之间,Agatston积分、体积积分及质量积分总体差异均有统计学意义(P均＜0.05)。随IR权重或DLIR级别增加,Agatston积分逐渐降低。以不同重建算法所获低、中、高密度斑块体积及管腔体积差异均有统计学意义(P均＜0.05);基于不同重建算法的平扫及增强图像中的主动脉CT值标准差差异均有统计学意义(P均＜0.05)。随IR权重增加或DLIR级别增高,噪声逐渐减小;DLIR-H图像噪声最低。结论 DLIR降低图像噪声能力优于迭代重建算法;重建算法对于Agatston积分、斑块体积及管腔体积等均有显著影响。 |
| 英文摘要: |
| Objective To explore the impact of reconstruction algorithm on the quantitative analysis of coronary artery CT. Methods Raw data of coronary artery CT, including calcium scoring scanning and angiography of 58 patients were collected. The images were reconstructed using adaptive statistical iterative reconstruction V (ASIR-V) with different blending weights (ASIR-V0%, ASIR-V50%, ASIR-V100%) and deep learning image reconstruction (DLIR) at low, medium and high levels (DLIR-L, DLIR-M, DLIR-H). Calcium scores (Agatston score, mass score and volume score), quantitative analysis results of plaques and lumen, CT value of pericoronary adipose tissue and of ascending aorta were measured and compared. Results Significant overall differences of Agatston score, volume score and mass score were found among different reconstruction algorithms (all P＜0.05). Agatston score gradually decreased with increasing IR weight or DLIR level. Significant differences of plaque volume and luminal volume were detected among low, medium and high-density plaques obtained with different reconstruction algorithms (all P＜0.05). For both non-enhanced and enhanced scanning, the reconstruction algorithm showed significant differences of standard deviation of aortic CT values were found among different reconstruction algorithms (both P＜0.05). With IR weight increased or DLIR level increased, the noise gradually decreased, and DLIR-H image had the lowest noise. Conclusion The ability of DLIR for reducing image noise was superior to that of iterative reconstruction algorithms. Reconstruction algorithm had significant impact on Agatston integral, plaque volume and luminal volume. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
