Study For Optimization Technology Of Multi-Pinhole Collimated X-ray Fluorescence Computed Tomography System

As a new molecular imaging mode,X-ray fluorescence CT(XFCT)is the production of X-ray fluorescence analysis(XRFA)technology and CT reconstruction technology.It could identify the concentration and distribution of elemental species by detecting XRF signals emitted by incident X-rays and reconstructing the fluorescence projection with specific algorithm,which plays an important role in biomedicine,material science,and earth science.Although XFCT based on X-ray tube source has become a research hotspot at present,problems such as excessive scattering noise seriously interfere with the effective detection of fluorescent signals,resulting in low signal-to-noise ratio of imaging systems,poor image quality,and difficulty in meeting detection limits.It is undoubtedly of significance to explore the distribution of scattered noise spectrum,accurately estimate the number of scattering noise,and analyze the relationship between the scattering noise and the detection angle for enhancing the imaging performance of XFCT imaging base on X-ray tube source.Thus,to derecase the scattering noises,this paper focuses on the optimization of the imaging performance of the multi-collimated XFCT based on X-ray tube source,and with Geant4,the following studies are carried out from the detection angle and reconstruction algorithm:(1)Studied for simulation and optimization of multi-collimated XFCT system based on Geant4.We studied the principle of multi-pinhole Collimated XFCT and simulated a polychromatic multi-pinhole XFCT system with Geant4 to research how the detection angle affects the signal-to-noise.By changing the tube voltage,the concentration of the GNPs and the diameter of the region of interest,the signal-to-noise ratio was calculated by simulating the detection process of emitted fluorescence and Compton scattering noise with different angles.The experimental results demonstrated that,the increase of the detection angle could effectively improve the signal-to-noise ratio.Furthermore,by changing the imaging method,phantom type and GNPs concentration,the projections at different detection angles were simulated and reconstructed with OSEM algorithm.Both visual effects and quantitative evaluation(CNR)indicated that detector angle optimization has more advantages for low-concentration imaging.(2)Studied for optimized EM-TV reconstruction algorithm based on scattering noise model.Due to the presence of scattered noise in the imaging,the MLEM reconstruction algorithm only iterates the fluorescent projection,which makes it difficult to eliminate statistical noise and the quality of the XFCT reconstructed image is poor.an optimized EM-TV algorithm based on the scattering noise model is proposed.The algorithm embeds the correction of scattered photons in the algorithm,estimates and iterates the fluorescence signal and scattering noise respectively to reduce the effect of the statistical noise;and the total variation is introduced TV as a penalty function to ensure the convergence of the reconstruction algorithm.We design and simulated the parallel multi-collimated XFCT imaging system with Geant4.The MLEM algorithm,optimized EM algorithm and optimized EM-TV algorithm were used to reconstruct the fluorescence projection and calculate the CNR value.The experimental results demonstrated that,optimized EM-TV algorithm could improve the accuracy of scattering noise estimation,greatly enhance the quality of the reconstructed image,and further improve the system performance of XFCT based on X-ray tube source.