Cargando…

A photoacoustic image reconstruction method using total variation and nonconvex optimization

BACKGROUND: In photoacoustic imaging (PAI), the reduction of scanning time is a major concern for PAI in practice. A popular strategy is to reconstruct the image from the sparse-view sampling data. However, the insufficient data leads to reconstruction quality deteriorating. Therefore, it is very im...

Descripción completa

Detalles Bibliográficos
Autores principales: Zhang, Chen, Zhang, Yan, Wang, Yuanyuan
Formato: Online Artículo Texto
Lenguaje:English
Publicado: BioMed Central 2014
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4148921/
https://www.ncbi.nlm.nih.gov/pubmed/25129644
http://dx.doi.org/10.1186/1475-925X-13-117
Descripción
Sumario:BACKGROUND: In photoacoustic imaging (PAI), the reduction of scanning time is a major concern for PAI in practice. A popular strategy is to reconstruct the image from the sparse-view sampling data. However, the insufficient data leads to reconstruction quality deteriorating. Therefore, it is very important to enhance the quality of the sparse-view reconstructed images. METHOD: In this paper, we proposed a joint total variation and L(p)-norm (TV-L(p)) based image reconstruction algorithm for PAI. In this algorithm, the reconstructed image is updated by calculating its total variation value and L(p)-norm value. Along with the iteration, an operator-splitting framework is utilized to reduce the computational cost and the Barzilai-Borwein step size selection method is adopted to obtain the faster convergence. RESULTS AND CONCLUSION: Through the numerical simulation, the proposed algorithm is validated and compared with other widely used PAI reconstruction algorithms. It is revealed in the simulation result that the proposed algorithm may be more accurate than the other algorithms. Moreover, the computational cost, the convergence, the robustness to noises and the tunable parameters of the algorithm are all discussed respectively. We also implement the TV-L(p) algorithm in the in-vitro experiments to verify its performance in practice. Through the numerical simulations and in-vitro experiments, it is demonstrated that the proposed algorithm enhances the quality of the reconstructed images with faster calculation speed and convergence.