Block matching sparsity regularization-based image reconstruction for incomplete projection data in computed tomography.

Block matching sparsity regularization-based image reconstruction for incomplete projection data in computed tomography.
Phys Med Biol. 2017 Nov 30;:
Authors: Cai A, Li L, Zheng Z, Zhang H, Wang L, Hu G, Yan B
Abstract
In medical imaging fields, many of the conventional regularization meth-ods, such as total variation or total generalized variation, impose strong prior assumptions on images which can only account for very limited classes of im-ages. A more reasonable sparse representation frame for images is still in strong need. Visually understandable images contain meaningful patterns and the combinations or collections of these patterns can be utilized to form some sparse and redundant representations which can promisingly facilitate image reconstructions. In this work, we propose and study the block matching spar-sity regularization (BMSR) and devise the optimization program using BMSR for computed tomography (CT) image reconstruction for incomplete projec-tion set. The program is built as a constrained optimization, minimizing the L1-norm of the coefficients of the image in the transformed domain subject to the data observation and positivity of image itself. To solve the program effi-ciently, a practical algorithm based on the proximal point algorithm is devel-oped and analyzed. In order to accelerate the convergence rate, a practical strategy for tuning the BMSR parameter is proposed and applied. The experi-mental results of ...

Source: Physics in Medicine and Biology - November 30, 2017 Category: Physics Authors: Cai A, Li L, Zheng Z, Zhang H, Wang L, Hu G, Yan B Tags: Phys Med Biol Source Type: research