非等中心锥束CT成像系统标定及三维重建

目的 研究非等中心锥束CT系统(CBCT)几何参数标定及投影图像重排, 提出新的方法以提高参数标定精度及三维(3D)重建质量, 以期突破现有算法要求等中心系统应用环境的限制。方法 针对非等中心CBCT系统建立坐标系, 推导标定模型中特定标志点空间坐标与其在投影图像中平面坐标之间的几何关系, 建立虚拟等中心CBCT系统, 提出面向非等中心CBCT系统的解析+迭代混合标定方法, 对投影图像进行重排, 在此基础上进行3D重建。将标定模型置于非等中心CBCT系统中, 获取各个旋转角度下的投影图像;提取投影图像中特定标志点坐标, 基于这些坐标数据, 采用所提方法计算虚拟等中心CBCT系统几何参数;并求取虚拟等中心锥束投影图像;最后采用FDK算法进行3D重建。结果 与文献[10]算法相比, 所提算法对转轴在成像板上的投影坐标(u^0' 和v^0' )标定精度相当, 而射线源到成像板的垂直距离(D')和射线源到转轴的距离(R')标定精度明显较优。结论 本研究混合标定算法可提高D'和R'的标定精度, 并可突破文献[10]算法等中心系统应用环境的限制。

Calibration and three-dimensional reconstruction for non-isocentric cone-beam CT imaging system

Objective Based on the research of geometric calibration and image rearrangement of non-isocentric cone-beam CT (CBCT) imaging system, this paper proposed a new calibration method, in order to improve calibration accuracy and three-dimensional (3D) reconstruction quality of non-isocentric CBCT system, and to break the restriction of existing algorithm which is only suitable for isocentric imaging system. Methods A coordinate system was established for non-isocentric CBCT imaging system. The geometric relationship between the spatial coordinates of the specific landmark and the corresponding projection coordinates in the imaging plane was deduced. An hybrid calibration approach of analytical and iterative methods was proposed. Based on the geometric parameters of the non-isocentric system, the projected images were rearranged and a virtual isocentric system was established. Through FDK algorithm, the parameters of the virtual isocentric system were used to reconstruct a 3D model. A calibration model was placed in the non-isocentric CBCT imaging system, and a sequence of projection images was acquired. The projection images were then processed to extract the specific landmark coordinates. The virtual isocentric CBCT system was established and then its imaging parameters were computed with the hybrid calibration approach on the basis of the landmark coordinates. The projected images were rearranged based on the parameters of the virtual isocentric system. Through the FDK algorithm, the calibration parameters were used to reconstruct a 3D model based on the rearranged projection images. Results The proposed hybrid approach was proved to be more accurate for distance from the source to the detector (D') and distance from the source to the rotation center (R') than literature [10] algorithm, while the calibration accuracy for position of the rotation axis on the detector in both directions (u₀' and v₀') were equal. Conclusion The proposed hybrid approach can be used to enhance accuracy of the calibration parameters D' and R' and overcome the limitation of the existing algorithm which was only suitable for isocentric imaging system.