Discrimination And Suppression Of The Backward-scattering Effect In Compton Imaging

Compton imaging is a promising technology for various applications including nuclear safety,nuclear medicine,and astrophysics.This technology relies on effective background suppression to ensure its sensitivity.But the backward-scattering effect(BSE)will increase imaging background,decrease signal-to-noise ratio(SNR),and degrade imaging precision.Therefore,discrimination and suppression of the BSE is significant for improving Compton imaging performance.Currently,there is only a few research work related to the BSE in the world,lacking of a deeper systemic study yet.In order to find the changing rule of the BSE affected by system structure,and illustrate the mechanism that how the BSE degrading imaging precision,we constructed a Compton imager based on double layers of CZT pixel array detectors using Geant4 Monte Carlo Package.Combined with numerical simulation,theoretical analysis and experimental verification,we improved a theoretical-calculation method for discrimination of backward-scattering imaging events(BSIEs),and proposed a novel algorithm to revive BSIEs for quasi-point sources.Finally,we realized the discrimination and suppression of the BSE,and improved the imaging performance of Compton imager.The research results show that high-Z-material front detector can increase the intrinsic imaging efficiency of Compton imager,but it will lead to serious BSE simultaneously.That is,the number of BSIEs can be equivalent to or even more than forward-scattering imaging events(FSIEs).However,BSIEs will be miss-reconstructed if taken as the default expected FSIEs,which make background increased and SNR decreased.Though properly reconstructed,the imaging precision of BSIEs is much worse than that of FSIEs.This is a universal phenomenon in Compton imager and it will degrade imaging performance.The higher the energy of the radioactive source,the further to the central axis of the detectors,the more serious the BSE.By theoretical analysis,we found the underlying mechanism why the imaging precision of BSIEs is much worse than that of FSIEs,i.e.,error propagation of energy resolution exacerbates the measurement uncertainty of large-angle scattering,resulting in the Compton cones of BSIEs much more discrete than that of FSIEs.After a long time of experimental measurement,the energy depositions of BSIEs in front and back detectors are expected to exhibit a concentration distribution.According to this feature,BSIEs can be effectively discriminated by energy threshold method.However,it is very difficult to get the concentrated energy ranges of BSIEs under conditions of rapid recognition or low counting.Based on the spatial geometrical layout among the radiation source and the two detectors,the angular range of BSIEs can be calculated theoretically.Then,combined with the detectors’ energy calibration curve,which can be calibrated experimentally,the energy ranges of BSIEs can be deduced in advance.Finally,the effective discrimination of BSIEs will be achieved even under conditions of low counting,and the discrimination efficiency can usually exceed 99%.Given the considerable number of BSIEs,the improvement of their imaging precision and effective utilization in image reconstruction will be helpful to increase the imaging efficiency of Compton imager.For commonly used quasipoint-source imaging applications,a novel algorithm named as self-adaption Compton imaging algorithm(SCIA)was proposed.This algorithm can significantly improve the imaging precision of Compton imager by iterative calculation,and realize the accurate identification and positioning of multiple quasi-point sources.Besides,SCIA is also effective to revive BSIEs for image reconstruction so that imaging efficiency can be nearly doubled.Through Geant4 simulation and experimental verification,the imaging performance of SCIA and traditional Compton imaging algorithm was comprehensively compared and evaluated,and the applicability of SCIA was verified in various scenarios.