Experimental Research of in Vivo Mouse Cardiac 4D Micro-CT Imaging Via Deformation Vector Field Registration
Xiaohui Gu,Guang Li,Ning Gu,Shouhua Luo
DOI: https://doi.org/10.1007/s11220-019-0246-y
2019-01-01
Sensing and Imaging
Abstract:In the radiation therapy, high-quality four-dimensional (4D) Cone-beam Computed Tomography data of patients with lung cancer can be obtained by deforming prior images using estimated deformation vector fields (DVF). In pre-clinical field, meanwhile, four-dimensional Micro-Computed Tomography (4D Micro-CT) imaging is also a promising tool to explore long-term cardiopulmonary diseases via in vivo mouse models. However, since the respiration and heart rates of mice are 8 times faster than those of humans, and the higher resolution of Micro-CT even requires more projection views, the 4D Micro-CT imaging of mice is more challenging than that of human being. In this paper, we integrate the DVF registration method with retrospective gating technology, and try to realize cardiac 4D Micro-CT imaging of in vivo mouse. The whole experimental research is completed on our self-developed Micro-CT platform, and the procedures of implementation are as follows. Firstly, the multi-turn projection data through a specific scanning protocol is obtained, and then the projection views locating in the respiratory resting period are picked and categorized into different groups in terms of phase information acquired from the respiratory and ECG gating sub-system. Secondly, the selected projection data is reconstructed via the ordered subsets simultaneous algebraic reconstruction technique algorithm as prior images. Finally, the images within different cardiac phases are made by deforming the prior images using optimized DVF. The results show every one 4D Micro-CT image series with 5 cardiac phases can be secured through one scan, which all bear clear heart contours and are with less streak artifacts, and the images made by the proposed method have higher signal noise ratio compared with those by conventional algorithms. A similar performance can be obtained when the angular sampling interval decreases. Hence, the proposed 4D cardiac imaging technology can effectively characterize the cardiac motion, and can be developed as a professional tool to image the heart of a living mouse dynamically.