4D‐CT Based CPD‐Demons Hybrid Image Registration for Pulmonary Functional Analysis

Research Square (Research Square)(2023)

引用 0|浏览3
暂无评分
摘要
Abstract Purpose : Pulmonary function analysis is essential for implementing functional avoidance in modern radiotherapy treatment planning. The gold standard for pulmonary function analysis is the SPECT V-Q scan. However, additional radiation exposure introduced and dynamic image fusion remains obstacles in treatment planning. Methods and Materials : This study proposes a hybrid deformable image registration method for pulmonary ventilation functional analysis. The proposed method separates the deformation of the human lung into two parts, deformation of the outer boundaries and the inner lung tissue. In reality, the deformation of the lung boundaries is constrained by adjacent organs such as the heart, parietal pleura, and diaphragm, where the lung boundaries exhibit non-linear, anisotropic deformation behavior, especially near the diaphragm; the inner lung tissue exhibit a more fluid-like, viscous deformation behavior. Therefore, in this study, the lung boundaries are modeled using the coherent point drift (CPD) registration method, while the inner lung tissue is registered using the Demons registration method. The proposed hybrid DIR method establishes the correspondence between exhalation and inhalation 3D-CT images of a 4D-CT data set. Finally, a ventilation functional map is then constructed based on the difference in Hounsfield Units (HU) and compared to the ventilation of SPECT. Results : For all 4D-CT SPECT data sets, the ventilation of the corresponding functional maps was found to deviate by no more than 7.92%. In addition, the lowest correlation between the corresponding functional maps is found to be 0.68, and the highest is 0.87. Conclusion : With the results of this study, the basic feasibility of the proposed DIR for pulmonary functional analysis is confirmed. Furthermore, the 4D-CT-based ventilation functional maps have higher resolution compared to SPECT ventilation maps and give evidence of the former as a replacement for the latter.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要