L{sub 1/2} regularization based numerical method for effective reconstruction of bioluminescence tomography
- School of Life Science and Technology, Xidian University, Xi'an 710071 (China)
Even though bioluminescence tomography (BLT) exhibits significant potential and wide applications in macroscopic imaging of small animals in vivo, the inverse reconstruction is still a tough problem that has plagued researchers in a related area. The ill-posedness of inverse reconstruction arises from insufficient measurements and modeling errors, so that the inverse reconstruction cannot be solved directly. In this study, an l{sub 1/2} regularization based numerical method was developed for effective reconstruction of BLT. In the method, the inverse reconstruction of BLT was constrained into an l{sub 1/2} regularization problem, and then the weighted interior-point algorithm (WIPA) was applied to solve the problem through transforming it into obtaining the solution of a series of l{sub 1} regularizers. The feasibility and effectiveness of the proposed method were demonstrated with numerical simulations on a digital mouse. Stability verification experiments further illustrated the robustness of the proposed method for different levels of Gaussian noise.
- OSTI ID:
- 22275547
- Journal Information:
- Journal of Applied Physics, Vol. 115, Issue 18; Other Information: (c) 2014 AIP Publishing LLC; Country of input: International Atomic Energy Agency (IAEA); ISSN 0021-8979
- Country of Publication:
- United States
- Language:
- English
Similar Records
TH-EF-207A-07: An Integrated X-Ray/bioluminescence Tomography System for Radiation Guidance and Tumor Evaluation
A multi-phase level set framework for source reconstruction in bioluminescence tomography