High Quality Cardiopulmonary Imaging Of In-vivo Micro-CT And Analysis Of Pulmonary Fibrosisof Small Animal Models

Micro-Computed Tomography(Micro-CT)is widely used in pre-clinical study because of its non-invasiveness and high resolution.When plenty of applications of Micro-CT imaging for small animal scanning,especially for moving organs such as the lung and the heart,are taking place,the existing problems are increasing prominently.All of those mainly manifest three aspects: firstly,in order to achieve high resolution imaging,the inherent problem of Micro-CT is its insufficient supply of light flux,which leads to lower signal noise ratio of reconstructed images.Moreover,due to the high respiration rates and heartbeats,the reconstructed images of small animals will be affected inevitablely by the motion artifacts.It is necessary to develop some corresponding hardware and software methods to improve the image quality.Secondly,there are few studies on the disease model of moving organs by micro-CT imaging,and how to make the Micro-CT imaging outcomes more representative is also an important issue.Finally,in longitudinal studies of disease models,micro-CT often produces a large number of image data.Due to the difference between the physiological structure of small animals and human beings,it is necessary to develop corresponding analysis methods in line with the characteristics of small animal micro-CT images.Based on the above reasons,this study focuses on three aspects: high-resolution imaging of moving organs,longitudinal imaging of disease models and post-processing analysis of micro-CT images.First of all,the method of in vivo cardiopulmonary imaging of small animal micro-CT is studied.In this paper,a retrospective respiratory gating system based on epipolar consistency is presented.The respiratory signal is extracted from the projection series by the value of the epipolar consistency metrics.On this basis,a new scanning scheme of multi-cycle multisegment speed-changing is realized.The respiratory signal is extracted without any additional hardware,and a better performance is achieved in reducing respiratory artifacts.Then a deformation vector field(DVF)driven cardiac imaging method is further studied.Through a specific scanning protocol,the multi-cycle projection data is obtained,and the projection views locating in the respiratory resting period are picked up and categorized into different groups.The aggregated projections are then reconstructed as the prior images by the ordered subsets simultaneous algebraic reconstruction technique(OS-SART)algorithm.By deforming these prior images using optimized DVF,the images within different cardiac phases are obtained.The results show 4D Micro-CT image series with 5 cardiac phases can be obtained through one scan,which all bear clear heart contours and are with less streak artifacts.Secondly,the mouse model of pulmonary fibrosis is given as the representative for the longitudinal studies of micro-CT imaging.Bleomycin-induced lung fibrosis was created by intratracheally instilling bleomycin into mice.Micro-CT imaging is performed to evaluate disease onset and progression at different time-points,coupled with histological analysis for the validation of the imaging results.Within two weeks after modeling,the pathological results show that the mice have varying degrees of acute inflammation,chronic inflammation and fibrosis,while in the micro-CT images,the different degrees of consolidation,floccus opacity,reticular opacity,ground-glass opacity,bronchiectasis and pneumothorax are observed.Comparing the micro-CT features with the pathological results in the same time-points,the causes of these features in micro-CT images are discussed.Finally,a specific quantitative analysis method is developed for the micro-CT images of pulmonary fibrosis mice.Preprocessing procedures such as image rotation,image smoothing,lung segmentation and airway segmentation are performed on 3D micro-CT image series.Then the extent of bronchial dilation,high density opacity and reticular opacity are quantitively assessed.The results of pulmonary fibrosis model mice are compared with those of healthy mice.The results show that there is no distinct difference between the results of quantitative analysis and the empirical scoring.