Mayo Clinic Rochester, 200 First Street SW, Rochester, Minnesota 55905, USA.
Med Phys. 2010 Feb;37(2):897-906. doi: 10.1118/1.3271110.
To determine the constancy of z-axis spatial resolution, CT number, image noise, and the potential for image artifacts for nonconstant velocity spiral CT data reconstructed using a flexibly weighted 3D filtered backprojection (WFBP) reconstruction algorithm.
A WFBP reconstruction algorithm was used to reconstruct stationary (axial, pitch=0), constant velocity spiral (pitch = 0.35-1.5) and nonconstant velocity spiral CT data acquired using a 128 x 0.6 mm acquisition mode (38.4 mm total detector length, z-flying focal spot technique), and a gantry rotation time of 0.30 s. Nonconstant velocity scans used the system's periodic spiral mode, where the table moved in and out of the gantry in a cyclical manner. For all scan types, the volume CTDI was 10 mGy. Measurements of CT number, image noise, and the slice sensitivity profile were made for all scan types as a function of the nominal slice width, table velocity, and position within the scan field of view. A thorax phantom was scanned using all modes and reconstructed transverse and coronal plane images were compared.
Negligible differences in slice thickness, CT number, noise, or artifacts were found between scan modes for data taken at two positions within the scan field of view. For nominal slices of 1.0-3.0 mm, FWHM values of the slice sensitivity profiles were essentially independent of the scan type. For periodic spiral scans, FWHM values measured at the center of the scan range were indistinguishable from those taken 5 mm from one end of the scan range. All CT numbers were within +/- 5 HU, and CT number and noise values were similar for all scan modes assessed. A slight increase in noise and artifact level was observed 5 mm from the start of the scan on the first pass of the periodic spiral. On subsequent passes, noise and artifact level in the transverse and coronal plane images were the same for all scan modes.
Nonconstant velocity periodic spiral scans can achieve z-axis spatial resolution, CT number accuracy, image noise and artifact level equivalent to those for stationary (axial), and constant velocity spiral scans. Thus, periodic spiral scans are expected to allow assessment of four-dimensional CT data for scan lengths greater than the detector width without sacrificing image quality.
确定使用灵活加权三维滤波反投影(WFBP)重建算法重建非恒定速度螺旋 CT 数据时,z 轴空间分辨率、CT 值、图像噪声和图像伪影的稳定性。
使用 WFBP 重建算法重建静止(轴向,螺距=0)、恒速螺旋(螺距=0.35-1.5)和非恒速螺旋 CT 数据,采集模式为 128x0.6mm(总探测器长度 38.4mm,z 飞行焦点技术),机架旋转时间为 0.30s。非恒速扫描使用系统的周期性螺旋模式,其中床以周期性方式进出机架。对于所有扫描类型,体积 CT DI 为 10mGy。对于所有扫描类型,作为名义切片宽度、床速和扫描视场位置的函数,测量 CT 值、图像噪声和切片灵敏度分布。使用所有模式对胸部体模进行扫描,并比较重建的横断和冠状平面图像。
在扫描视场的两个位置采集的数据中,不同扫描模式之间的切片厚度、CT 值、噪声或伪影差异可以忽略不计。对于名义切片厚度为 1.0-3.0mm 的切片,切片灵敏度分布的 FWHM 值基本与扫描类型无关。对于周期性螺旋扫描,在扫描范围中心测量的 FWHM 值与在扫描范围一端测量的 5mm 处的 FWHM 值无法区分。所有 CT 值均在 +/-5HU 范围内,所有扫描模式的 CT 值和噪声值均相似。在周期性螺旋扫描的第一圈,在扫描开始处 5mm 处观察到噪声和伪影水平略有增加。在后续的通过中,所有扫描模式的横断和冠状平面图像的噪声和伪影水平相同。
非恒定速度周期性螺旋扫描可以实现与静止(轴向)和恒速螺旋扫描相同的 z 轴空间分辨率、CT 值准确性、图像噪声和伪影水平。因此,周期性螺旋扫描有望在不牺牲图像质量的情况下,评估大于探测器宽度的扫描长度的四维 CT 数据。