CAJ | 학술논문

目的：获取基于4D-CT和变形图像配准方法，建立肺通气功能的三维分布图像。方法在患者自由呼吸状态下进行肺部4D-CT图像采集和重建；用B样条三维变形配准算法，对4D-CT中不同相位的两个CT图像系列进行配准，获得三维变形场文件，再将该文件转换为雅可比行列式矩阵，根据其物理含义进行量化分析，得到任意横断面的通气分布图像。将此图像伪彩化后与CT图像融合，再进行任意冠状面、矢状面重建，从而得到直观的横断面、冠状面、矢状面三维方向的通气功能分布。将自动勾画并计算出的不同功能肺的体积，与吸气末和呼气末肺体积之差进行相关分析。结果实现了快速获得直观的肺通气功能的三维分布。结果表明，不同功能肺体积与呼吸过程中肺体积的最大变化显著相关（P<0.05）。结论基于4D-CT和变形图像配准来获取肺通气功能的三维分布图像简便、快捷，具有可行性。
목적：획취기우4D-CT화변형도상배준방법，건립폐통기공능적삼유분포도상。방법재환자자유호흡상태하진행폐부4D-CT도상채집화중건；용B양조삼유변형배준산법，대4D-CT중불동상위적량개CT도상계렬진행배준，획득삼유변형장문건，재장해문건전환위아가비행렬식구진，근거기물리함의진행양화분석，득도임의횡단면적통기분포도상。장차도상위채화후여CT도상융합，재진행임의관상면、시상면중건，종이득도직관적횡단면、관상면、시상면삼유방향적통기공능분포。장자동구화병계산출적불동공능폐적체적，여흡기말화호기말폐체적지차진행상관분석。결과실현료쾌속획득직관적폐통기공능적삼유분포。결과표명，불동공능폐체적여호흡과정중폐체적적최대변화현저상관（P<0.05）。결론기우4D-CT화변형도상배준래획취폐통기공능적삼유분포도상간편、쾌첩，구유가행성。
Objective To obtain 3D distribution images of pulmonary ventilation based on 4D-CT and deformation image registration (DIR). Methods The 4D-CT data sets were acquired with patients in free-breathing. 3D displacement vector ifeld (DVF) of two different phase 4D-CT image pairs was calculated out by using 3D B-spline DIR algorithms, which was converted to Jacobian determinant. Then, the axial section of grayscale ventilation was obtained by quantitative analysis of the determinant. The pseudo-color was put on the axial grayscale ventilation before fused with the reference CT images. And based on the axial ventilation, the coronal and sagittal sections of ventilations can be reconstructed. Thus, 3D visualization ventilations have been implemented. The contours of the ventilation regions with different Jacobian values were delineated and the volumes of them were calculated. Results Based on 4D-CT images of patients in free-breathing and multi-resolution 3D B-spline DIR, the 3D visualization ventilation can be easily obtained and quantiifed. The volume change of lung is signiifcantly related with the volume of different ventilation regions at level 0.05 (bilateral). Conclusion It is simple, convenient and feasible to obtain the 3D distribution of pulmonary ventilation based on the 4D-CT images and 3D B-spline DIR.