Department of Nuclear Medicine and Molecular Imaging, University Medical Center Groningen, University of Groningen, 9700 RB Groningen, The Netherlands.
Eur J Nucl Med Mol Imaging. 2011 Dec;38(12):2257-68. doi: 10.1007/s00259-011-1900-3. Epub 2011 Aug 18.
Metallic prosthetic replacements, such as hip or knee implants, are known to cause strong streaking artefacts in CT images. These artefacts likely induce over- or underestimation of the activity concentration near the metallic implants when applying CT-based attenuation correction of positron emission tomography (PET) images. Since this degrades the diagnostic quality of the images, metal artefact reduction (MAR) prior to attenuation correction is required.
The proposed MAR method, referred to as virtual sinogram-based technique, replaces the projection bins of the sinogram that are influenced by metallic implants by a 2-D Clough-Tocher cubic interpolation scheme performed in an irregular grid, called Delaunay triangulated grid. To assess the performance of the proposed method, a physical phantom and 30 clinical PET/CT studies including hip prostheses were used. The results were compared to the method implemented on the Siemens Biograph mCT PET/CT scanner.
Both phantom and clinical studies revealed that the proposed method performs equally well as the Siemens MAR method in the regions corresponding to bright streaking artefacts and the artefact-free regions. However, in regions corresponding to dark streaking artefacts, the Siemens method does not seem to appropriately correct the tracer uptake while the proposed method consistently increased the uptake in the underestimated regions, thus bringing it to the expected level. This observation is corroborated by the experimental phantom study which demonstrates that the proposed method approaches the true activity concentration more closely.
The proposed MAR method allows more accurate CT-based attenuation correction of PET images and prevents misinterpretation of tracer uptake, which might be biased owing to the propagation of bright and dark streaking artefacts from CT images to the PET data following the attenuation correction procedure.
金属假体置换物,如髋关节或膝关节植入物,已知会在 CT 图像中产生强烈的条纹伪影。这些伪影可能会导致在应用基于 CT 的正电子发射断层扫描(PET)图像衰减校正时,高估或低估金属植入物附近的活性浓度。由于这会降低图像的诊断质量,因此需要在衰减校正前进行金属伪影减少(MAR)。
所提出的 MAR 方法,称为基于虚拟正弦图的技术,通过在称为 Delaunay 三角剖分网格的不规则网格中执行的二维 Clough-Tocher 三次插值方案,用 2-D 替换受金属植入物影响的正弦图的投影箱。为了评估所提出方法的性能,使用物理体模和 30 项包括髋关节假体的临床 PET/CT 研究。将结果与在西门子 Biograph mCT PET/CT 扫描仪上实现的方法进行了比较。
体模和临床研究均表明,在所对应的亮条纹伪影和无伪影区域,该方法与西门子 MAR 方法的性能相当。然而,在所对应的暗条纹伪影区域,西门子方法似乎不能正确校正示踪剂摄取,而所提出的方法始终会增加被低估区域的摄取量,从而使其达到预期水平。这项观察结果得到了实验体模研究的证实,该研究表明,该方法更接近真实的活动浓度。
所提出的 MAR 方法允许更准确地基于 CT 的 PET 图像衰减校正,并防止由于在衰减校正过程后,从 CT 图像传播到 PET 数据的亮条纹和暗条纹伪影而导致示踪剂摄取的错误解释。