Quantitative cone-beam x-ray luminescence computed tomography with 3D TV denoising based on Split Bregman method

Abstract

Cone-beam X-ray luminescence computed tomography (CB-XLCT) is a noninvasive molecular imaging technique that reflects the distribution of fluorescent nanomaterials in the imaged object. It is urgent to describe the quantitative relationship between the reconstruction and the concentration of the fluorescent nanomaterials. However, in the field of CB-XLCT, most researches aim to improve the imaging accuracy, ignoring further quantitative evaluation of the reconstruction intensity. In this work, the quantitative evaluation for CB-XLCT is studied. In addition, to improve the quantitative performance, a new strategy based on fast iterative shrinkage-thresholding algorithm (FISTA) and 3D Total- Variation (TV) denoising with Split Bregman (SB) method (FISTA-TV) is proposed for CB-XLCT reconstruction. In FISTA-TV, FISTA is applied to get a L1-regularized sparse reconstruction in CB-XLCT and the Split Bregman method is used to solve the TV denoising problem. With the FISTA-TV strategy, the sparse results yielded by FISTA together with 3D TV denoising based on Split Bregman, alleviate the illness of the inverse problem of CB-XLCT, making the relationship between the reconstruction intensity and the actual concentration of fluorescent nanomaterials more accurate. Computer simulations have shown the quantitative reconstruction and evaluation for CB-XLCT is improved with the proposed FISTA-TV algorithm, compared to Algebraic Reconstruction Technique (ART), Tikhonov regularization, FISTA.