摘要
目的通过检测MRI的空间几何失真程度大小,研究其直接用于放疗计划设计的可行性。方法自制油脂管按每15°间隔辐射状排列置于泡沫模体内,制成模体1;自制油脂管按2.7cm间距网格状排列置于泡沫模体内,制成模体2。以Marconi1.5T超导型磁共振仪采用头部FSET2WI序列,分别对2个模体行平行于辐射状和网格状油脂管冠位面扫描和垂直于上述各面的横断位扫描,然后进行冠状面DRR重建。将获得的图像分为不同大小的5个视窗范围,测量直接扫描图像和DRR图像的边界绝对位置误差并计算其几何畸变率。结果对直接扫描的MRI,在11cm×11cm的视窗范围内,其边界最大位置误差基本为零;在22cm×22cm视窗内,其边界最大位置误差仅为1.1mm;在27.5cm×27.5cm的视窗范围内,其边界最大位置误差为1.4mm。对冠状面DRR重建图像,在16.5cm×16.5cm的视窗范围内,其边界最大位置误差<1.0mm;在22cm×22cm视窗范围内,其边界最大位置误差为1.4mm;在27.5cm×27.5cm的视窗范围内,其边界最大位置误差为2.0mm。结论对高场强的磁共振仪,在假定组织为均匀密度的前提下,MRI可以直接用于颅内肿瘤的放疗计划设计。
Objective Objective To study the feasibility of direct applying MRI to TPS by measuring the geometric distortion of MRI. Methods Phantom 1 was made by marking radical arrangement of fat pipelines every 15° in foam phantom and Phantom 2 by net arrangement at every 2.7?cm. Coronal and transverse scan parallel with the radial and net fat pipeline in the phantom was performed at sequence of cranial FSE T2WI by Marconi 1.5T superconduct MRI machine. Geometric distortion in 5 different fields of view (FOV) was calculated by measuring the absolute borderline error of direct scan and coronal digital reconstructed radiograph (DRR). Results When FOV was equal or smaller than 11.0?cm×11.0?cm, 22.0?cm×22.0?cm and 27.5?cm×27.5?cm, the geometric distortion was 0.0, 1.1 and 1.4?mm of direct scanning MRI and 1.0,1.4 and 2.0?mm of coronal DRR. Conclusion When the magnetic field of MR is high and the tissue shows uniform density, MRI can be applied on TPS directly.
出处
《中华放射肿瘤学杂志》
CSCD
北大核心
2005年第4期327-329,共3页
Chinese Journal of Radiation Oncology
关键词
MRI
颅内肿瘤
计划设计
计算机辅助
Magnetic resonance imaging
Radiotherapy planning,computed-asisted
Feasibility studies
