胸部CT扫描中迭代去金属伪影算法减少金属伪影的有效性

发布时间：2018-10-17 14:37

【摘要】：目的评价胸部CT扫描中迭代去金属伪影算法(IMAR)减少金属伪影的有效性。方法采用仿真胸部体模(仿真体模组),在T5椎体两侧各植入1枚椎弓根螺钉。以130kV管电压和自动毫安秒(care dose 4D)技术,在植入螺钉前后行胸部CT扫描。对无金属图像进行滤波反投影(FBP)算法重建,对有金属螺钉的图像分别采用FBP算法和IMAR重建,选取螺钉邻近层面的主动脉、肺、椎旁软组织为ROI,测量噪声值和CT值,并计算CT值相对于无金属植入物图像CT值的偏移程度(ΔHU)。另回顾性分析我院26例接受胸部CT检查、且扫描范围内有椎弓根螺钉植入物的患者(病例组),扫描条件、图像重建方式参数同体模扫描,测量金属伪影邻近脊椎骨组织及背部软组织CT值,并由2名放射科医师分别对两种重建图像进行主观图像质量评价。结果植入螺钉后,仿真体模组主动脉、肺、椎旁软组织IMAR重建图像的噪声值低于FBP重建图像(P均0.05),IMAR重建的ΔHU明显小于FBP重建图像(P均0.01)。病例组FBP重建和IMAR重建方式的脊椎骨组织CT值和背部软组织CT值差异均有统计学意义(P均0.05)。对FBP重建和IMAR重建图像的主观评分差异有统计学意义(P0.05)。结论 IMAR算法可明显减少条状伪影,并可校正受影响的组织CT值,使之更接近于无金属情况下的真实值。
[Abstract]:Objective to evaluate the effectiveness of iterative de-metal artifact algorithm (IMAR) in chest CT scanning. Methods one pedicle screw was implanted on both sides of T _ 5 vertebra by emulation chest phantom (emulation body model group). Chest CT scanning was performed before and after screw implantation using 130kV tube voltage and automatic millisecond (care dose 4 D technique. The non-metal images were reconstructed by filtered back-projection (FBP) algorithm. The images with metal screws were reconstructed by FBP algorithm and IMAR algorithm respectively. The aorta and lung adjacent to the screws were selected to measure the noise value and CT value of the adjacent soft tissue of the vertebrae for ROI,. The deviation of CT value relative to the CT value of non-metallic implants image was calculated (螖 HU). In addition, 26 cases of thoracic CT with pedicle screw implants were analyzed retrospectively. The CT values of the adjacent vertebrae and back soft tissue of metal artifact were measured, and the subjective image quality of the two reconstructed images was evaluated by two radiologists. Results after screw implantation, IMAR images of aorta, lung and paravertebral soft tissue in the simulated model group were lower than those in FBP reconstruction image (P 0.05), IMAR), and 螖 HU was significantly lower than FBP reconstruction image (P0.01). There were significant differences in CT value of vertebral tissue and CT value of back soft tissue between FBP reconstruction and IMAR reconstruction in the case group (all P 0. 05). The subjective scores of FBP and IMAR images were significantly different (P0.05). Conclusion the IMAR algorithm can significantly reduce the strip artifacts and correct the CT value of the affected tissues, which is closer to the real value in the metal-free case.
【作者单位】：重庆医科大学附属第一医院放射科;西门子(中国)医疗事业部;
【基金】：国家临床重点专科建设项目([2013]544)
【分类号】：R816.4

【参考文献】