Multi-energy Spectral CT Material Recognition Based On Photon Counting Detector

The multi-spectral CT system based on photon counting detector can identify and count photons with energy in a certain range by setting threshold values,and obtain projection data under multiple energy channels,fully utilizing the energy dependence of X-ray attenuation.The identification and separation of different substances has great advantages in medical diagnosis,safety monitoring,industrial production and other fields.In this paper,the material identification of multi-spectral CT is studied based on photon counting detector.The main research work is as follows:(1)Using the photon counting detection technology and the K-edge feature of matter,collecting the projections under multiple energy channels,deducing the linear model of projection decomposition based on the energy spectrum CT projection basis decomposition model,and solving it by least squares method.The material identification of the multi-energy spectrum CT contrast agent was realized,and the concentration of the contrast agent was quantitatively analyzed.(2)Proposed a method based on the Beetle Antennae Search Algorithm(BAS)to minimize the total variation(TV)and applied to pluripotent Spectral CT reconstruction.On this basis,the principal component analysis method is used to extract the energy spectrum information of the reconstructed image under multiple energy channels,and the model relationship between the principal component and the color channel RGB is established.The material identification and color characterization of multi-energy spectrum CT are realized.