Image Reconstruction Of Inverse Geometry Micro-CT Based On Array Micro-focal X-Ray Source

Micro computed tomography(Micro-CT)is a kind of micro imaging technology using X-ray to obtain the internal microstructure of samples.It can directly display,measure and analyze the 3D microstructure of samples without damaging the samples.It has been widely used in medical,industrial,archaeological,materials,national defense and military,bioengineering and other fields.It plays an important role in the detection of complex parts,biological tissues and internal defects of products and the accurate measurement of structural dimensions.In order to meet the needs of different applications,there are many types of micro CT systems or scanning imaging methods in recent years,such as inverse geometry CT(IGCT)composed of distributed multifocal ray source and small high-resolution detector.Compared with the traditional micro CT system,IGCT has the advantages of large field of view(FOV),less cone beam artifacts,less scattering effect and high signal-to-noise ratio.However,there are some shortcomings in IGCT system,such as projection data truncation and local data redundancy.The traditional analytical image reconstruction algorithms such as filtered back projection(FBP)are not suitable for IGCT system.In this paper,the image reconstruction of the inverse geometry micro CT(IG-Micro-CT)system based on the electron beam scanning array microfocal ray source and small area array high-resolution radiation detector is studied(1)The geometric model of IG-Micro-CT is established,the parameter relationship and formula of IG-Micro-CT are analyzed and deduced,and the simulation projection data of IG-Micro-CT is obtained by programming.Aiming at the characteristics of IG-Micro-CT projection data truncation and local data redundancy,an iterative reconstruction algorithm with prior constraints is proposed and tested by simulation experiments.(2)In order to meet the requirements of high detection efficiency applications,IG-Micro-CT system adopts multi focus fast sparse scanning mode.In view of the characteristics of less projection and incomplete data in this sparse scanning mode,the total variation regularization algorithm is added to the iterative process of image reconstruction to constrain the sparsity of the reconstructed image.In order to get the best balance between the image quality and the detection efficiency,the Shepp Logan phantom is used to optimize the parameters such as the number of focus and the number of scanning divisions to determine the optimal parameters.Under the optimal parameters,the simulation results of SIRT + TVM,SART + TVM,MART + TVM,MLEM + TVM,CGLS + TVM are compared and the convergence of the algorithm is tested,The three-dimensional reconstruction of IG-Micro-CT is simulated by SART and CGLS algorithms using three-dimensional Shepp Logan phantom.(3)Several groups of IG-Micro CT scanning experiments were completed on the IG-Micro-CT experimental platform,and high-quality IG-Micro-CT test images were obtained.The reconstructed images of different focus number scanning experiments and different algorithms were compared,and the 3D rendering of reconstruction results was displayed.