Chang Zheng, Cai Jing, Wang Ziheng, Yin Fang-Fang
Department of Radiation Oncology, Duke University, Durham, North Carolina, USA.
J Radiosurg SBRT. 2011;1(2):109-115.
The objective of this study is to investigate accuracy of motion tracking of cine magnetic resonance imaging (MRI) for image-guided stereotactic body radiotherapy. A phantom platform was developed in this work to fulfill the goal. The motion phantom consisted of a platform, a solid thread, a motor and a control system that can simulate motion in various modes. To validate its reproducibility, the phantom platform was setup three times and imaged with fluoroscopy using an electronic portal imaging device (EPID) for the same motion profile. After the validation test, the phantom platform was evaluated using cine MRI at 2.5 frames/second on a 1.5T GE scanner using five different artificial profiles and five patient profiles. The above profiles were again measured with EPID fluoroscopy and used as references. Discrepancies between measured profiles from cine MRI and EPID were quantified using root-mean-square (RMS) and standard deviation (SD). Pearson's product moment correlational analysis was used to test correlation. The standard deviation for the reproducibility test was 0.28 mm. The discrepancies (RMS) between all profiles measured by cine MRI and EPID fluoroscopy ranged from 0.30 to 0.49 mm for artificial profiles and ranged from 0.75 to 0.91 mm for five patient profiles. The cine MRI sequence could precisely track phantom motion and the proposed motion phantom was feasible to evaluate cine MRI accuracy.
本研究的目的是调查用于图像引导立体定向体部放射治疗的电影磁共振成像（MRI）运动跟踪的准确性。在这项工作中开发了一个体模平台以实现该目标。运动体模由一个平台、一根实心线、一个电机和一个能够模拟各种模式运动的控制系统组成。为了验证其可重复性，将体模平台设置三次，并使用电子门静脉成像设备（EPID）通过透视成像获取相同的运动轮廓。在验证测试之后，在1.5T通用电气扫描仪上以每秒2.5帧的速度使用电影MRI对体模平台进行评估，使用五种不同的人工轮廓和五种患者轮廓。上述轮廓再次通过EPID透视测量并用作参考。使用均方根（RMS）和标准差（SD）对电影MRI测量的轮廓与EPID之间的差异进行量化。使用皮尔逊积矩相关分析来测试相关性。可重复性测试的标准差为0.28毫米。电影MRI和EPID透视测量的所有轮廓之间的差异（RMS）对于人工轮廓范围为0.30至0.49毫米，对于五种患者轮廓范围为0.75至0.91毫米。电影MRI序列可以精确跟踪体模运动，并且所提出的运动体模对于评估电影MRI准确性是可行的。
