Li Baojun, Tang Xiangyang
Department of Radiology, Boston University School of Medicine, Boston, Massachusetts, USA.
Department of Radiology, Emory University School of Medicine, Atlanta, Georgia, USA.
J Appl Clin Med Phys. 2025 Sep;26(9):e70220. doi: 10.1002/acm2.70220.
Ultra-wide coverage CT (> 128 detector rows) makes it possible to image a heart or brain in a single rotation but are associated with large cone angles, which can severely degrade the image quality.
This study evaluate the image quality and artifact levels of a dual-focal-spot single-detector (DFSSD) CT geometry designed to achieve 140 mm z-axis coverage, through a simulation study.
The DFSSD CT system employs two x-ray focal spots spaced 90 mm apart along the z-axis and a 100 mm CT detector. It provides 140 mm z-axis coverage at the iso-center. A research CT simulation package (CatSim) was employed throughout this study. To compare the image quality, three system geometries were simulated: (A) single focal spot with 40 mm detector (VCT40), (B) single focal spot with 140 mm detector (VCT140), and (C) the DFSSD CT system. A simulated helical body phantom (45 cm x 30 cm x 16 cm) containing various bony structures was used to assess cone-beam artifacts and noise uniformity across the edge, central, and intermediate slices under different acquisition modes: axial half scan, axial full scan, and helical scan at various helical pitch (0.5, 0.75, 1.0). All images were reconstructed using a cone-beam filtered back-projection algorithm with 3D cone-angle-dependent pixel-wise weighting.
Under helical scan conditions (pitch 0.5 to 1.0), all geometries demonstrated similar levels of cone-beam artifacts and noises. Slightly increased artifacts observed on some slices in VCT140 and DFSSD CT, but overall image quality are acceptable. However, under axial half scan condition, the VCT140 geometry exhibited significantly worse artifacts and image uniformity in intermediate and edge slices compared to VCT40, whereas DFSSD CT showed artifact and image uniformity comparable to VCT40.
Simulation results indicate that the DFSSD CT geometry can achieve 140 mm z-axis coverage while maintaining image quality similar to the VCT40 system.
超宽覆盖CT（探测器排数>128排）能够在单次旋转中对心脏或大脑进行成像，但会产生较大的锥角，这可能会严重降低图像质量。
本研究通过模拟研究评估一种设计用于实现140mm z轴覆盖的双焦点单探测器（DFSSD）CT几何结构的图像质量和伪影水平。
DFSSD CT系统采用两个沿z轴间隔90mm的x射线焦点和一个100mm的CT探测器。它在等中心处提供140mm的z轴覆盖范围。在本研究中使用了一个研究CT模拟软件包（CatSim）。为了比较图像质量，模拟了三种系统几何结构：（A）配备40mm探测器的单焦点（VCT40），（B）配备140mm探测器的单焦点（VCT140），以及（C）DFSSD CT系统。使用一个包含各种骨质结构的模拟螺旋体模（45cm×30cm×16cm）来评估在不同采集模式下（轴向半扫描、轴向全扫描以及不同螺旋 pitch（0.5、0.75、1.0）的螺旋扫描）边缘、中心和中间层面的锥束伪影和噪声均匀性。所有图像均使用带有3D锥角相关逐像素加权的锥束滤波反投影算法进行重建。
在螺旋扫描条件下（pitch为0.5至1.0），所有几何结构的锥束伪影和噪声水平相似。在VCT140和DFSSD CT的一些层面上观察到伪影略有增加，但总体图像质量是可以接受的。然而，在轴向半扫描条件下，与VCT40相比，VCT140几何结构在中间和边缘层面表现出明显更差的伪影和图像均匀性，而DFSSD CT显示出与VCT40相当的伪影和图像均匀性。
模拟结果表明，DFSSD CT几何结构能够实现140mm的z轴覆盖，同时保持与VCT40系统相似的图像质量。
