文献速递心血管CT

光子计数CT冠状动脉CT血管造影的标准与超高分辨率协议:定性与定量图像质量分析

Qualitative and quantitative image quality of coronary CT angiography using photon-counting computed tomography: Standard and Ultra-high resolution protocols.
PubMed 文献 · Eur J Radiol · 2024-03-12

研究纳入45例患者(标准采集29例、超高分辨率采集16例),在双源光子计数CT上比较不同层厚、卷积核与QIR级别对冠状动脉CT血管造影图像质量的影响;定量显示标准分辨率下0.6 mm、Bv40、QIR4的CNR最高(25.8±4.1),超高分辨率下0.4 mm、Bv40、QIR4的CNR最高(21.5±3.9);定性以Bv44卷积核联合QIR 3/4主观质量最佳。

要点

  • 标准分辨率下增大层厚使CNR平均提高12.5%,更锐利的卷积核使CNR平均降低8.7%。
  • 标准分辨率最高CNR见于0.6 mm、Bv40、QIR4图像(25.8±4.1)。
  • 超高分辨率下QIR 2-4、Bv56、0.2 mm层厚可保持远端血管显示。

核心结论

  • 光子计数CT在冠状动脉与支架评估中表现出较高的定性与定量图像质量。

研究设计与样本

  • 对45例患者(标准采集29例、超高分辨率采集16例)行双源光子计数CT冠状动脉CT血管造影,分别应用较平滑与较锐利的卷积核设置进行图像重建分析。

关键影像表现/征象

  • 标准分辨率下,较大层厚使CNR平均提高12.5%,更锐利的卷积核使CNR平均降低8.7%。
  • 标准分辨率最高CNR见于0.6 mm、Bv40、QIR4图像:25.8±4.1,最低见于0.4 mm、Bv44、QIR0图像:8.3±1.6(p < 0.001)。
  • 超高分辨率下最高CNR见于0.4 mm、Bv40、QIR4图像:21.5±3.9,最低见于0.2 mm、Bv56、QIR0图像:3.6±0.8(p < 0.001)。
  • 标准重建两种层厚下,Bv44卷积核联合QIR级别3/4的图像主观质量最高;超高分辨率0.2 mm层厚下Bv56联合QIR4的主观图像质量最高。
  • 使用QIR 2-4、Bv56及0.2 mm层厚可保持远端血管显示。

诊断标准与关键数值

  • 定性图像质量采用4点Likert量表评估。
  • 图像噪声以SD表示,并计算近段与远段的SNR和CNR。
  • 标准采集参数为0.4/0.6 mm层厚、Bv40/Bv44卷积核、QIR级别0-4。
  • 超高分辨率采集参数为0.2/0.4 mm层厚、Bv44/Bv56卷积核、QIR级别0-4。

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

PURPOSE: We aimed to identify the optimal reconstruction settings based on qualitative and quantitative image quality parameters on standard and ultra-high resolution (UHR) images using photon-counting CT (PCCT).

METHOD: We analysed 45 patients, 29 with standard and 16 with UHR acquisition, applying both smoother and sharper kernel settings. Coronary CT angiography images were performed on a dual-source PCCT system using standard (0.4/0.6 mm slice thickness, Bv40/Bv44 kernels, QIR levels 0-4) or UHR acquisition (0.2/0.4 mm slice thickness, Bv44/Bv56 kernels, QIR levels 0-4). Qualitative image quality was assessed using a 4-point Likert scale. Image noise (SD), signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated in both the proximal and distal segments.

RESULTS: On standard resolution, larger slice thickness resulted in an average increase of 12.5 % in CNR, whereas sharper kernel led to an average 8.7 % decrease in CNR. Highest CNR was measured on 0.6 mm, Bv40, QIR4 images and lowest on 0.4 mm, Bv44, QIR0 images: 25.8 ± 4.1vs.8.3 ± 1.6 (p < 0.001). On UHR images, highest CNR was observed on 0.4 mm, Bv40, QIR4 and lowest on 0.2 mm, Bv56 and QIR0 images: 21.5 ± 3.9vs.3.6 ± 0.8 (p < 0.001). Highest qualitative image quality was found on images with Bv44 kernel and QIR level 3/4 with both slice thicknesses on standard reconstruction. Additionally, Bv56 with QIR4 on 0.2 mm slice thickness images showed highest subjective image quality. Preserved distal vessel visualization was detected using QIR 2-4, Bv56 and 0.2 mm slice thickness.

CONCLUSIONS: Photon-counting CT demonstrated high qualitative and quantitative image quality for the assessment of coronaries and stents.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Vattay B, Boussoussou M, Vecsey-Nagy M, et al. Qualitative and quantitative image quality of coronary CT angiography using photon-counting computed tomography: Standard and Ultra-high resolution protocols. Eur J Radiol. 2024. PMID: 38493558. doi:10.1016/j.ejrad.2024.111426

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。