Imaging and Medical Physics, Department of Radiology and Imaging Sciences, Emory University School of Medicine, 1701 Uppergate Dr., C-5018, Atlanta, GA, 30322, USA.
Med Phys. 2017 Dec;44(12):6239-6250. doi: 10.1002/mp.12618. Epub 2017 Nov 1.
We present the methodology for analyzing and optimizing the sampling structure of projection data acquisition in axial multidetector CT (MDCT) and cone beam CT (CBCT) under the framework of sampling on lattice. Specifically, we propose and evaluate the scheme of interlaced detector cell binning for suppression of longitudinal aliasing artifacts. In addition, we investigate the proposed scheme's capability of mitigating shift variation in spatial resolution and possibility of improving CB image reconstruction accuracy.
Under the framework of sampling on lattice, the proposed scheme is evaluated using an axial MDCT with its architecture similar to that of state-of-the-art CT scanners for diagnostic imaging in the clinic. The widely used FDK algorithm is adopted for image reconstruction, in which either horizontal/latitudinal or vertical/longitudinal interpolation is used for lining-up of projection data between interlaced detector cells. Using a spiral clock phantom, the capability of suppressing aliasing artifacts and possibility of improving reconstruction accuracy is quantitatively investigated. The in-plane spatial resolution, as assessed by the modulation transfer function (MTF), and its shift-variant property are quantitatively assessed using wire phantoms, while the through-plane spatial resolution and its shift-variant behavior are assessed by the slice sensitivity profile (SSP) using thin foil phantoms.
The preliminary results show that the interlaced detector cell binning can suppress longitudinal aliasing artifacts effectively, while the shift variation in spatial resolution and reconstruction inaccuracy can be mitigated moderately. In addition, the direction, along which the interpolation is carried out to line up projection data between the interlaced detector cells for image reconstruction, plays a significant role in determining the in-plane and through-plane spatial resolution.
The scheme of interlaced detector cell binning with longitudinal interpolation for data lining-up is an effective solution for suppression of longitudinal aliasing artifacts in axial MDCT and CBCT.
我们提出了在晶格采样框架下分析和优化轴向多探测器 CT（MDCT）和锥形束 CT（CBCT）投影数据采集采样结构的方法。具体来说，我们提出并评估了交错探测器单元 binning 的方案，以抑制纵向混叠伪影。此外，我们研究了该方案在减轻空间分辨率变化和提高 CB 图像重建准确性方面的能力。
在晶格采样框架下，使用与临床诊断成像用的最先进 CT 扫描仪结构相似的轴向 MDCT 评估所提出的方案。采用广泛使用的 FDK 算法进行图像重建，其中在交错探测器单元之间的投影数据中使用水平/横向或垂直/纵向插值进行对齐。使用螺旋时钟体模，定量研究了抑制混叠伪影的能力和提高重建准确性的可能性。使用线模体评估调制传递函数（MTF）评估的面内空间分辨率及其位移变化特性，而使用薄箔模体评估切片灵敏度分布（SSP）评估的穿透平面空间分辨率及其位移变化行为。
初步结果表明，交错探测器单元 binning 可以有效抑制纵向混叠伪影，同时可以适度减轻空间分辨率的位移变化和重建不准确。此外，在交错探测器单元之间进行数据对齐以进行图像重建的插值方向在确定面内和穿透平面空间分辨率方面起着重要作用。
具有纵向插值的数据对齐交错探测器单元 binning 方案是抑制轴向 MDCT 和 CBCT 中纵向混叠伪影的有效解决方案。
