Application Research On Basis Material Decomposition Of Spectral CT

Since the late 1970s,the exploration of spectral CT has never stopped.From traditional CT that can not discern X-ray energy to dual-energy CT(DECT)that discriminates two X-ray energy spectra,and from DECT to real spectral CT that can distinguish multiple energy channels.Spectral CT uses transmission data under different X-ray energies passing through the measured object to provide richer image information than traditional CT.It has huge application potential in many hot areas such as improving soft tissue contrast,low-dose imaging and tissue quantitative analysis.Dual energy CT is the most widely implemented form of spectral CT.Compared with traditional CT,dual energy CT has the advantage of removing the beam hardening effect and obtaining multi-parameter information including the electron density and effective atomic number.In this paper,a numerical simulation for double material decomposition of dual-energy CT is firstly completed.This method adopts the strategy of directly solving the projection function equation to obtain the decomposition coefficient projections,furtherly,multiple parameter images can be reconstructed.Due to the extremely complex composition of the actual scanned object,it is difficult to achieve further accurate material distinction for the double material decomposition.In order to adapt to the real scanning situation and get more component imaging,this paper proposes a DECT multi-material decomposition algorithm based on pseudoinverse and optimal solution criterion,recorded as MMD-POSC algorithm.For the ill-posed problem that using dual-energy CT to solve the density images of more than two components,the MMD-POSC algorithm converts it into a well-posed problem by combining the mass and volume conservation constraints.Moreover,the MMD-POSC algorithm adopts the idea of solving pixel by pixel,and selects a set of optimal solutions for each pixel from the material library according to the optimal basis selection criteria.Finally,using the basis material decomposition coefficients to reconstruct the multiple images about different parameter distribution information of the scanned object.Both numerical simulation experiment and actual data experiment have confirmed the effectiveness of the algorithmThe double material decomposition based on the basis material decomposition model can accurately describe the linear attenuation coefficient of the material in the absence of the K absorption edge.However,for some material,there is a K-edge discontinuity in the linear attenuation coefficient.The linear attenuation coefficient originally described by the linear combination of the photoelectric cross section and the Compton cross section,is not enough to accurately describe the attenuation characteristics of the material.Taking into account elements whose linear attenuation coefficient has K-edge discontinuities,this paper proposes a numerical simulation algorithm using three-material decomposition for K-edge spectral CT(KCT).In the selection of energy spectral,considering the K-edge effect,it is necessary to select appropriate energy spectral values.In the projection decomposition,the projection value of the decomposition coefficient of the basis material is still obtained by direct calculation.Furtherly,the MMD-POSC algorithm of dual-energy CT is extended to the case of spectral CT with K-edge effect,that is,proposing a multi-material decomposition algorithm based on KCT,denoted as KMMD-POSC algorithm.Experiments shows that:when there exists K-edge effect,the KMMD-POSC algorithm of K-edge spectral CT is obviously superior to the MMD-POSC algorithm of dual-energy CT.