Design Of Small Animal CT Software System With Helical Cone Beam

CT system has been playing an important role in clinical diagnosis,industrial testing and preclinical research since it was officially launched in the early 1970 s.Due to the demand for high-precision CT imaging system in the field of preclinical research and the development of micro-focus X-ray source,flat panel detector and other technologies,special CT imaging system with high spatial resolution,high sensitivity and high scanning speed developed for small animal experiments has a wide market prospect.At present,most small animal CT systems use axial cone beam scanning,which has limitations in the imaging field,and artifacts are serious when the cone angle is large.However,the spiral CT system widely used in clinical diagnosis can carry out fast and accurate imaging of long objects.Therefore,this paper will discuss how to design and develop a more practical spiral cone beam CT system for small animals with spiral cone beam scanning mode.Based on the existing cone-beam small animal CT hardware equipment in the laboratory,this paper designed and developed a small animal CT software system combining the helical cone beam scanning mode and wireless data transmission.It solves the the problem that traditional cone-beam small animal CT is difficult to image long objects such as big rats,and improves the applicability of the system.There is no need to adjust the hardware structure of the system,two scanning modes can be used only by selecting the software algorithm.In addition,data transmission based on Wi Fi makes the system get rid of the dependence on high-cost photoelectric slip ring and reduces the system cost.The main contribution of this paper:(1)Using CUDA,the GPU parallel acceleration algorithm of Katsevich 3D precise filtering backprojection reconstruction algorithm used in the system is designed,which solves the problem of long reconstruction time of Katsevich algorithm and improves the applicability of the system in practical research;(2)An interval endpoint smoothing processing formula for Katsevich reconstruction algorithm is proposed,which can be used to deal with the cone-beam covering method,and improved quantitative;(3)A dictionary learning improved scattering correction algorithm based on image postprocessing is proposed for image scattering correction,and the scattering correction results are good.The Shepp-Logan phantom was used to simulate the system test,and the reconstruction results were consistent with the real data.QRM micro-CT high spatial resolution volume model was used for testing,and the result could distinguish 100 μm line pairs.Finally,the system is used to image rats and other objects,and the imaging results are good.