Rehfeld Niklas S, Heismann Björn J, Kupferschläger Jürgen, Aschoff Philip, Christ Gunter, Pfannenberg Anna C, Pichler Bernd J
Laboratory for Preclinical Imaging and Imaging Technology, Werner Siemens-Foundation, University of Tübingen, Röntgenweg 13, D-72076 Tübingen, Germany.
Med Phys. 2008 May;35(5):1959-69. doi: 10.1118/1.2903476.
In present positron emission tomography (PET)/computed tomography (CT) scanners, PET attenuation correction is performed by relying on the information given by a single CT scan. The scaling of the linear attenuation coefficients from CT x-ray energy to PET 511 keV gamma energy is prone to errors especially in the presence of CT contrast agents. Attenuation correction based upon two CT scans at different energies but performed at the same time and patient position should reduce such errors and therefore improve the accuracy of the reconstructed PET images at the cost of introduced additional noise. Such CT scans could be provided by future PET/CT scanners that have either dual source CT or energy sensitive CT. Three different dual energy scaling methods for attenuation correction are introduced and assessed by measurements with a modified NEMA 1994 phantom with different CT contrast agent concentrations. The scaling is achieved by differentiating between (1) Compton and photoelectric effect, (2) atomic number and density, or (3) water-bone and water-iodine scaling schemes. The scaling method (3) is called hybrid dual energy computed tomography attenuation correction (hybrid DECTAC). All three dual energy scaling methods lead to a reduction of contrast agent artifacts with respect to single energy scaling. The hybrid DECTAC method resulted in PET images with the weakest artifacts. Both, the hybrid DECTAC and Compton/photoelectric effect scaling resulted also in images with the lowest PET background variability. Atomic number/density scaling and Compton/photoelectric effect scaling had problems to correctly scale water, hybrid DECTAC scaling and single energy scaling to correctly scale Teflon. Atomic number/density scaling and hybrid DECTAC could be generalized to reduce these problems.
在目前的正电子发射断层扫描(PET)/计算机断层扫描(CT)扫描仪中,PET衰减校正依靠单次CT扫描提供的信息来进行。从CT X射线能量到PET 511 keVγ能量的线性衰减系数的缩放容易出错,尤其是在存在CT造影剂的情况下。基于在不同能量但在相同时间和患者位置进行的两次CT扫描的衰减校正应能减少此类误差,从而以引入额外噪声为代价提高重建PET图像的准确性。这样的CT扫描可由未来具有双源CT或能量敏感CT的PET/CT扫描仪提供。本文介绍了三种不同的用于衰减校正的双能量缩放方法,并通过使用具有不同CT造影剂浓度的改良NEMA 1994体模进行测量来评估。通过区分(1)康普顿效应和光电效应、(2)原子序数和密度或(3)水-骨和水-碘缩放方案来实现缩放。缩放方法(3)称为混合双能量计算机断层扫描衰减校正(hybrid DECTAC)。所有三种双能量缩放方法相对于单能量缩放都能减少造影剂伪影。混合DECTAC方法产生的PET图像伪影最弱。混合DECTAC和康普顿/光电效应缩放都还产生了PET背景变异性最低的图像。原子序数/密度缩放和康普顿/光电效应缩放难以正确缩放水,混合DECTAC缩放和单能量缩放难以正确缩放聚四氟乙烯。原子序数/密度缩放和混合DECTAC可以进行推广以减少这些问题。