A level set method for Bayesian tomographic reconstruction

Shiying Zhao1
1Department of Mathematics and Computer Science, University of Missouri, Saint Louis, USA

Tóm tắt

We present a level-set method for Bayesian tomographic reconstruction. A novel image prior is derived from the mean curvature evolution of level sets of an image. As it has been studied in image processing with nonlinear diffusion, this prior encourages the stabilization of an edge while the reconstructed image is smoothed along both sides of the edge. An algorithm of iterated coordinate decent was implemented with the proposed prior using Brent's method for one-dimensional optimization. Our simulation results demonstrated that our algorithm can outperform existing priors for preserving edges during tomographic reconstruction without introducing additional artifacts.

Từ khóa

#Level set #Bayesian methods #Tomography #Image reconstruction #Kinetic energy #Image processing #Optimization methods #Image restoration #Mathematics #Computer science

Tài liệu tham khảo

10.1007/BF00375127 10.1109/83.491321 zhao, 2002, A mechanical image model for Bayesian tomographic reconstruction, Preprint 10.1109/ICIP.2000.899510 jonsson, 1998, Total variation regularization in positron emission tomography, Reports 98-48 U C L A Computational and Applied Mathematics lee, 1995, Bayesian image reconstruction in SPECT using higher order mechanical models as priors, IEEE Trans Med Imag, 4, 669 yu, 2001, Edge-preserving tomographic reconstruction with nonlocal regularization, Preprint frese, 2001, Adaptive wavelet graph model for Bayesian tomographic reconstruction, Preprint 10.1016/0167-2789(92)90242-F 10.1016/0021-9991(88)90002-2 10.1109/83.236536 leahy, 2000, Recent developments in iterative image reconstruction for PET and SPECT, IEEE Trans Med Imag, 19, 257, 10.1109/TMI.2000.848177 10.1137/0729052