实用老年医学

实用老年医学 ›› 2025, Vol. 39 ›› Issue (12): 1195-1201.doi: 10.3969/j.issn.1003-9198.2025.12.002

• 专题研究 • 上一篇    下一篇

基于混合效应模型的光子计数冠状动脉CT血管成像图像质量参数分析

徐晶, 周逸驹, 尹佳妮, 梁静, 余鸿鸣, 陈文萍, 李辉, 牡丹, 薛秋苍   

  1. 210008 江苏省南京市,南京医科大学鼓楼临床医学院医学影像科(徐晶,牡丹,薛秋苍);
    210008 江苏省南京市,南京大学医学院附属鼓楼医院医学影像科(周逸驹,尹佳妮,梁静,余鸿鸣,陈文萍,李辉,薛秋苍)
  • 收稿日期:2025-09-12 发布日期:2025-12-26
  • 通讯作者: 薛秋苍,Email:295507338@qq.com
  • 基金资助:
    国家自然科学基金面上项目(82272065);第15批中国博士后特别资助项目(2022T150317);江苏省卫健委医学科研面上项目(M2022066);南京市医学科技发展重点项目(ZKX23019);南京鼓楼医院新技术发展项目(XJSFZJJ202026,XJSFZLX202114);南京鼓楼医院2024年临床研究专项资助项目(2024-LCYJ-PY-80)

Analysis of image quality parameters of photon-counting coronary CT angiography based on mixed-effects model

XU Jing, ZHOU Yiju, YIN Jiani, LIANG Jing, YU Hongming, CHEN Wenping, LI Hui, MU Dan, XUE Qiucang   

  1. Department of Radiology, Nanjing Drum Tower Hospital Clinical College of Nanjing Medical University, Nanjing 210008, China (XU Jing, MU Dan, XUE Qiucang);
    Department of Radiology, Nanjing Drum Tower Hospital, the Affiliated Hospital of Nanjing University Medical School, Nanjing 210008, China (ZHOU Yiju, YIN Jiani, LIANG Jing, YU Hongming, CHEN Wenping, LI Hui, XUE Qiucang)
  • Received:2025-09-12 Published:2025-12-26
  • Contact: XUE Qiucang, Email:295507338@qq.com

PDF (PC)

27

摘要： 目的 探讨光子计数CT(PCCT)在冠状动脉CT血管成像(CCTA)中不同层厚及卷积核组合对图像质量的影响,明确最优成像参数。 方法 前瞻性纳入2025年6—8月于南京大学附属鼓楼医院因疑似冠心病(CAD)而行PCCT-CCTA检查的40例患者,分别采用Bv40、Bv48、Bv56三种卷积核与0.4 mm、0.6 mm两种层厚组合重建图像。在主动脉根部及冠状动脉管腔测量图像CT值、噪声、信噪比(SNR)、对比噪声比(CNR)及峰值信噪比(PSNR)等指标。采用线性混合效应模型分析各参数对图像质量的主效应及交互效应,结合箱线图和相关分析,评估图像稳定性和指标间的关系。 结果 薄层图像与锐利卷积核可提升边缘清晰度,但同时显著增加图像噪声并降低图像稳定性。Bv56 + 0.4 mm组合噪声最高,SNR和CNR最低。Bv40 + 0.6 mm组合在图像噪声、SNR、CNR、PSNR等多个指标上表现最佳,箱线图显示其四分位距最小,参数波动性最低。卷积核是影响图像质量的最主要因素(均P<0.01),卷积核与层厚间未发现显著交互效应。 结论 在PCCT-CCTA中,Bv40卷积核结合0.6 mm层厚可在维持空间分辨率的同时最大程度降低图像噪声,提升整体图像质量与稳定性。

关键词: 光子计数CT, 图像质量, 层厚, 卷积核

Abstract: Objective To investigate the impact of different slice thicknesses and kernels on image quality of coronary CT angiography (CCTA) using photon-counting CT (PCCT), and to identify the optimal reconstruction parameters. Methods A total of 40 patients with suspected coronary artery disease (CAD) who underwent CCTA at Nanjing Drum Tower Hospital from June to August 2025 were prospectively enrolled. Images were reconstructed using three kernels (Bv40, Bv48, Bv56) and two slice thicknesses (0.4 mm, 0.6 mm). Quantitative measurements of CT attenuation, image noise, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and peak signal-to-noise ratio (PSNR) were performed at the aortic root and coronary arteries. A linear mixed-effects model was used to analyze the main and interaction effects of the kernel and slice thickness. Box plots and correlation analysis were conducted to assess image stability and parameter relationships. Results Although thin-slice reconstruction combined with sharper kernels enhanced edge sharpness, it significantly increased image noise and reduced image stability. The Bv56 + 0.4 mm combination yielded the highest noise level and the lowest SNR and CNR. The Bv40 + 0.6 mm combination demonstrated the best performance across multiple metrics, including image noise, SNR, CNR, and PSNR. Box plots showed this parameter combination had the smallest interquartile ranges, indicating minimal parameter fluctuation.The kernel type was the most dominant factor affecting all image quality metrics (all P<0.01).No significant interaction effect was found between the kernel and slice thickness. Conclusions For CCTA using PCCT, the Bv40 kernel + 0.6 mm slice thickness provides an optimal balance between spatial resolution and noise control, achieving superior image quality and consistency.

Key words: photon-counting CT, image quality, kernel, slice thickness

中图分类号: