Linear adaptive noise-reduction filters for tomographic imaging: Optimizing for minimum mean square error
- Univ. of California, Berkeley, CA (United States)
This thesis solves the problem of finding the optimal linear noise-reduction filter for linear tomographic image reconstruction. The optimization is data dependent and results in minimizing the mean-square error of the reconstructed image. The error is defined as the difference between the result and the best possible reconstruction. Applications for the optimal filter include reconstructions of positron emission tomographic (PET), X-ray computed tomographic, single-photon emission tomographic, and nuclear magnetic resonance imaging. Using high resolution PET as an example, the optimal filter is derived and presented for the convolution backprojection, Moore-Penrose pseudoinverse, and the natural-pixel basis set reconstruction methods. Simulations and experimental results are presented for the convolution backprojection method.
- Research Organization:
- Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
- Sponsoring Organization:
- Department of Health and Human Services
- DOE Contract Number:
- AC03-76SF00098
- OSTI ID:
- 10148667
- Report Number(s):
- LBL--34376; ON: DE94011345; CNN: Grant HL 07367
- Country of Publication:
- United States
- Language:
- English
Similar Records
Constrained least squares projection filtering: A new technique for the reconstruction of emission computed tomographic images
A numerical investigation of Tomographic filters
Data acquisition errors and the choice of filter in filtered backprojection algorithms
Conference
·
Sat Jan 31 23:00:00 EST 1987
· IEEE Trans. Nucl. Sci.; (United States)
·
OSTI ID:6700104
A numerical investigation of Tomographic filters
Conference
·
Thu Dec 31 23:00:00 EST 1992
· Transactions of the American Nuclear Society; (United States)
·
OSTI ID:5861982
Data acquisition errors and the choice of filter in filtered backprojection algorithms
Conference
·
Thu Jun 01 00:00:00 EDT 1989
·
OSTI ID:5970006