Title :
Wavelet based methods for multiscale tomographic reconstruction
Author :
Bhatia, M. ; Karl, W.C. ; Willsky, A.S.
Author_Institution :
Lab. for Inf. & Decision Syst., MIT, Cambridge, MA, USA
Abstract :
The authors combine wavelet bases with a natural-pixel type representation of an object, originally developed for incomplete data tomography problems, to construct “natural wavelet bases” for tomographic reconstruction. For the case of noisy projection data the authors use this multiscale framework to explicitly account for noise by calculating multiscale MAP estimates based on a certain self-similar prior
Keywords :
computerised tomography; image reconstruction; medical image processing; wavelet transforms; incomplete data tomography problems; multiscale MAP estimates; multiscale framework; multiscale tomographic reconstruction; natural wavelet bases; natural-pixel type representation; noisy projection data; self-similar prior; wavelet based methods; Discrete wavelet transforms; Equations; Image reconstruction; Reconstruction algorithms; Sparse matrices; Stacking; Tomography; Vectors; Wavelet coefficients; Wavelet transforms;
Conference_Titel :
Engineering in Medicine and Biology Society, 1994. Engineering Advances: New Opportunities for Biomedical Engineers. Proceedings of the 16th Annual International Conference of the IEEE
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-2050-6
DOI :
10.1109/IEMBS.1994.412121
