基于四维CT技术肺内孤立性肿瘤位移模型的建立与验证

目的基于四维CT(4DCT)测量肺内孤立性肿瘤三维方向的位移,建立肿瘤的位移模型并进行验证。方法对建模样本中290例肺内孤立性肿瘤行4DCT模拟定位和主动呼吸控制(ABC)下螺旋扫描,并基于4DCT测量不同肺段肿瘤三维方向位移,在ABC图像上构建静止状态下肿瘤体积。用模拟定位机测量患侧膈肌的运动幅度,用肺功能仪测量肺活量及潮气量。采集患者性别、年龄、身高、体重、呼吸频率,肿瘤所在肺叶、肺段等信息。采用多元线性回归对肿瘤三维方向的位移与患者性别、年龄、身高、体重、呼吸频率、肿瘤体积、位置、膈肌运动幅度等因素进行分析,建立肺内孤立性肿瘤三维方向的位移模型。序贯收集17例肺内孤立性肿瘤对位移模型的计算结果进行测试。结果上叶肺内孤立性肿瘤三维方向的位移模型分别为X上=-0.267+0.002TV+0.446DM,Y上=-1.704+0.004TV+0.725DM+2.250SⅡ+1.349SⅢ,Z上=0.043+0.626DM+0.599SⅡ+0.519SⅢ。中叶及上、下舌段肿瘤三维方向的位移模型分别为X中=0.539+0.758DM,Y中=-2.316+2.707DM+0.009TV,Z中=0.717+1.112DM。下叶肿瘤三维方向的位移模型分别为X下=-0.425+0.004TV+0.857DM,Y下=4.691+4.817DM+0.005TV-0.307RR+3.148SⅨ+2.655SⅩ,Z下=0.177+0.003TV+0.908DM。(DM:膈肌运动幅度,TV:潮气量,RR:呼吸频率,SⅡ:后段,SⅢ:前段,SⅨ:外侧底段,SⅩ:后底段)。位移模型的预测结果与4DCT法实测肿瘤的位移相近(P>0.05)。结论肺内孤立性肿瘤三维方向位移的主要影响因素为膈肌运动幅度和患者潮气量,同一肺叶不同肺段肿瘤头脚方向的位移差异显著,中叶不同肺段之间肿瘤的位移相近。位移模型法能较好地预测肺内孤立性肿瘤的位移,为个体化靶区构建提供参考。

Establishment and verification of a displacement model for the solitary pulmonary lesion based on 4D-CT technology

Objective To measure the displacement of the solitary pulmonary lesion (SPL) located in different pulmonary segments based on 4D-CT technology (4DCT) and to establish and verify a relevant mathematical model of tumor displacement. Methods The modeling samples of 290 SPLs were subject to both 4DCT and active breath control (ABC) spiral CT scans. The tumor displacement in different pulmonary segments was measured based on 4DCT images. The tumor volume was obtained by contouring the gross tumor volume (GTV) on ABC spiral CT images. The diaphragm movement was measured by X-ray simulator. The vital capacity and tidal volume were gained by pneumatometer. The baseline data, such as gender, age, height, weight, respiratory rate, and tumor lobe and segment where tumors were located, were collected. Multivariate linear regression was used to analyze the correlation between the 3D-tumor displacement and gender, age, height, weight, respiratory rate, tumor location, volume and diaphragm movement. The displacement model was established based on the modeling sample of 290 cases. Then,it was verified by comparing the tumor displacement derived from the model with that of 4DCT technology based on the randomly selected 17 SPLs. Results The displacement model for tumors located in the upper lobe was established as Xup=-0.267+0.002TV+0.446DM, Yup=-1.704+0.004TV+0.725DM+2.250SII+1.349SIII and Zup=0.043+0.626DM+0.599SII+0.519SIII. The displacement model of the middle lobe tumors was Xmid=0.539+0.758DM, Ymid=-2.316+2.707DM+0.009TV and Zmid=0.717＋1.112DM. The displacement model for tumors located in the lower lobe was Xlow=-0.425+0.004TV+0.857DM,Ylow=4.691+4.817DM+0.005TV-0.307RR+3.148SIX+2.655SX and Zlow=0.177+0.003TV+0.908DM.(DM:diaphragm movement, TV:tidal volume, RR:respiratory rate, SII:posterior segment, SIII:anterior segment, SIX:lateral basal segment, SX:posterior basal segment). There was no significant difference between two results derived from the displacement model and 4DCT technology.(P>0.05). Conclusions The diaphragm movement and tidal volume are the main influencing factors of 3D lung tumor displacement. The tumor displacement in the superior-inferior direction is correlated with different pulmonary segments of the upper and lower lobes. The displacement of tumors located in different segments of middle lobes is similar. The displacement model can predict the displacement of SPLs located in different lobes, providing reference for individualized delineation of PTV.