Reconstruction Image Quality Analysis With A Stationary Multi-beam X-ray Digital Breast Tomosynthesis

Breast cancer is one of the most common and deadly cancers.Related studies have proven that early detection and treatment of breast cancer can effectively improve the survival rate of patients.Examination of breast cancer generally uses mammography.However,traditional mammography provides a two-dimensional image,which results in overlapping images and a high recall rate.With the development of related technologies,Digital Breast Tomosynthesis(DBT)technology has been proven to be able to reconstruct several projection images obtained from limited angles using reconstruction algorithms to improve the detection rate of breast cancer and reduce recalls.rate.The DBT system that has been commercialized at present is a single X-ray source.Multi-angle projection through mechanical movement is extremely likely to cause motion artifacts.The fixed multi-beam X-ray light source distribution is adopted,and the projection image acquisition is completed by emitting X-rays one by one,which can eliminate motion artifacts and reduce the projection image acquisition time.First,this paper proposes seven different X-ray source distributions for the distribution of X-ray sources in a fixed multi-beam X-ray mammography system,and simulates seven different geometric configurations on the GATE(Geant4 Application for Tomographic Emission)platform X-ray source configurations,each configuration contains 25 X-ray sources.The classic back-projection(BP)algorithm and its two variants(?-trimmed-BP and PCA-BP)were used to reconstruct the acquired projection images.The results are analyzed and compared in the spatial and frequency domains.The experimental results show that of the seven geometric configurations of the X-ray source,the rectangular non-uniform configuration performs best.Shows better performance in detail.In addition,in the analysis of the three algorithms in the frequency domain by the modulation transfer function,?-trimmed-BP also shows better image quality.Secondly,the Monte Carlo approach was carried out through the GATE simulation platform,and the structure of a fixed multi-beam equal-angle linear digital mammography system was simulated to obtain 15 projection images.It uses three iterative algorithms to reconstruct it from 1-20 times.For three iterative reconstruction algorithms,including Simultaneous Algebraic Reconstruction Technique(SART),Maximum-Likelihood Expectation Maximization(MLEM),Ordered-Subset MLEM,OS-MLEM)reconstructed images with different iterations were evaluated using the Artifact Spread Function(ASF),Modulation Transfer Function(MTF),and Noise Power Spectrum(NPS).Considering the performance of reconstructing images with different iterations in eliminating out-of-focus plane artifacts,suppressing noise,and reconstructing the sharpness of the object in the frequency domain,the experimental results show that the algorithm's effect deteriorates sharply when the number of iterations is greater than 8.At 4,6,and 3,the reconstructed image quality of SART,MLEM,and OS-MLEM is generally optimal.