A Maximum a Posteriori Probability Expectation Maximization Algorithm for Image Reconstruction in Emission Tomography

Abstract

The expectation maximization method for maximum likelihood image reconstruction in emission tomography, based on the Poisson distribution of the statistically independent components of the image and measurement vectors, is extended to a maximum aposteriori image reconstruction using a multivariate Gaussian a priori probability distribution of the image vector. The approach is equivalent to a penalized maximum likelihood estimation with a special choice of the penalty function. The expectation maximization method is applied to find the a posteriori probability maximizer. A simple iterative formula is derived for a penalty function that is a weighted sum of the squared deviations of image vector components from their a priori mean values. The method is demonstrated to be superior to pure likelihood maximization, in that the penalty function prevents the occurrence of irregular high amplitude patterns in the image with a large number of iterations (the so-called "checkerboard effect" or "noise artifact").

Smoothing Parameter Estimation for Markov Random Field Classification of non-Gaussian Distribution Image

“…In this research, we applied expectation-maximization (EM) algorithm (Levitan and Herman, 1987) to compute the MAP. In Equation (2), U(Y | L) and U(L) denote spectral and spatial energy terms, respectively.…”

“…In this equation, c is the class label of a given pixel (i) of the category and s is the class label of surrounding pixel s. Due to the MRF neighboring concept which says that 1) a site cannot be a neighbor with itself i / ∈ N i ; and 2) the neighborhood relationship is mutual (i ∈ N j ⇐⇒ j ∈ N i ) (Levitan and Herman, 1987), CLCMC is converted to GCLCMC to show the global spatial frequency distribution of each pair of classes in the image:…”

Multigrid tomographic inversion with variable resolution data and image spaces

“…where p(y|x) is the probability density of Y given x, and S(x) is a stabilizing function designed to regularize the inversion [68], [69]. If S(x) = − log p(x), where p(x) is the image prior probability density, this results in the maximum a posteriori (MAP) estimate of x.…”

Application of Discrete Tomographic Methods in Nondestructive Testing and Materials Science

“…• Probabilistic optimization: maximum a posteriori expectation maximization (MAP EM) [86,131], expectation maximization (EM) [55], maximum likelihood (ML) [178].…”