A True-color CT Imaging Method Based On Energy Spectrum Filtering Separation

With the development of technology, the traditional X-ray CT imaging technology can’tmeet the modern industry needs of functional imaging for component distinguish andidentification. This is due to the inconsistency of X-ray source and reconstruction algorithm.In the current CT systems, X-ray spectrum is continuous with different energy segmentswhich is produced by X-ray source, and the attenuation coefficient of object is varied with theX-ray energy. But the current reconstruction algorithm, either analytical or iterative, are basedon the assumption: X-ray is single ray. This inconsistency is easy to result a phenomenon ofhardening artifacts, and make the CT reconstruction poor, and can’t distinguish the materialcomponent when their densities are similar. To solve this problem, researchers havesuccessfully developed a kind of photon-counting detector, which can discriminate the nergyof X-ray incident photon. This technique implement spectrum separation during theprojection acquisition process, however, the technology is not yet mature, and has somelimitations in the practical engineering application.In order to expand the existing CT imaging system and realize distinction of materialcomponent, this paper researchs a true-color CT imaging method based on energy spectrumfiltering separation in view of photon counting imaging detector. Firstly, spectrum filtering isrealized and monochromatic projection is gotten by filtering at X-ray source. Based on somepriori information and the spectrum filtering separation conditions, several parameters, suchas the number of separation times, voltage and range of filtering, are obtained. So that we canobtain a different spectrum segments projection sequence which can be approximate to singleenergy after the filtering separation. The simulation results show that: this imaging method isfeasible. Then a soft-threshold filtering TV-OSEM algorithm is researched at the condition oflow dose because of filtering. And the feasibility of this reconstruction algorithm is verified by simulation and experiment. What’s more, PCA analysis is used to fusion the reconstructionimages, and three principal components is corresponded to RGB component in color space,so we can get a true-color CT image which can reflect the material component information.Finally, three simulations show that: for some similar density of the detect objects, thismethod can distinguish the material components. Therefore, a true-color CT imaging methodbased on energy spectrum filtering separation can meet the requirement of componentdiscrimination, so that we can identify the material component further.