Mutaf Y D, Antolak J A, Brinkmann D H
Radiation Oncology Department, Mayo Clinic, 200 1st Street SW, Rochester, Minnesota 55905, USA.
Med Phys. 2007 May;34(5):1615-22. doi: 10.1118/1.2717404.
Accurate delineation of target volumes is one of the critical components contributing to the success of image-guided radiotherapy treatments and several imaging modalities are employed to increase the accuracy in target identification. Four-dimensional (4D) techniques are incorporated into existing radiation imaging techniques like computed tomography (CT) to account for the mobility of the target volumes. However, these methods in some cases introduce further inaccuracies in the target delineation when further quality assurance measures are not implemented. A source of commonly observed inaccuracy is the misidentification of the respiration cycles and resulting respiration phase assignments used in the construction of the 4D patient model. The aim of this work is to emphasize the importance of optimal respiration phase assignment during the 4DCT image acquisition process and to perform a quantitative assessment of the effect of inaccurate phase assignments on the overall image quality. The accuracy of the phase assignment was assessed by comparison with an independent calculation of the respiration phases. Misplaced phase assignments manifest themselves as deformations and artifacts in reconstructed images. These effects are quantified as volumetric discrepancies in the localization of target objects represented by spherical phantoms. Measurements are performed using a fully programmable motion phantom designed and built at Mayo Clinic (Rochester, MN). Implementation of a case based independent check and correction procedure is also demonstrated with emphasis on the use of this procedure in the clinical environment. Review of clinical 4D scans performed in this institution showed discrepancies in the phase assignments in about 40% of the cases when compared to our independent calculations. It is concluded that for improved image reconstruction, an independent check of the sorting procedure should be performed for each clinical 4DCT case.
精确勾勒靶区体积是图像引导放射治疗成功的关键因素之一，目前采用了多种成像方式来提高靶区识别的准确性。四维（4D）技术被纳入现有的放射成像技术，如计算机断层扫描（CT），以考虑靶区体积的移动性。然而，在某些情况下，如果没有进一步实施质量保证措施，这些方法会在靶区勾勒中引入更多不准确性。常见的不准确性来源之一是在构建4D患者模型时对呼吸周期的错误识别以及由此产生的呼吸相位分配。这项工作的目的是强调在4DCT图像采集过程中最佳呼吸相位分配的重要性，并对不准确的相位分配对整体图像质量的影响进行定量评估。通过与呼吸相位的独立计算进行比较来评估相位分配的准确性。错误的相位分配在重建图像中表现为变形和伪影。这些影响被量化为球形体模所代表的靶物体定位中的体积差异。使用梅奥诊所（明尼苏达州罗切斯特）设计和制造的完全可编程运动体模进行测量。还展示了基于病例的独立检查和校正程序的实施，重点是该程序在临床环境中的应用。对该机构进行的临床4D扫描的回顾显示，与我们的独立计算相比，约40%的病例在相位分配上存在差异。得出的结论是，为了改善图像重建，应对每个临床4DCT病例的排序程序进行独立检查。
