Assessing quantification for the EMS algorithm

Abstract A framework is developed for assessing the potential of the EMS (estimate, maximize, and smooth) algorithm to correctly quantify the emission it aims to reconstruct. Maximal eigenvector localization bounds of Perron-Frobenius theory are used to show that the ability of the EMS algorithm to quantify differences in emission is fundamentally limited by the smoothing imposed and the properties of the underlying model matrix in the algorithm. In particular, it is established that, for nonnegative irreducible smoothing, the EMS algorithm will always fail to correctly reconstruct zero emission intensity.

[1]  P. Green.  On Use of the EM Algorithm for Penalized Likelihood Estimation , 1990 .

[2]  D L Bailey,et al.  SPECT evaluation of arterial perfusion in regional chemotherapy. , 1988, Journal of nuclear medicine : official publication, Society of Nuclear Medicine.

[3]  J. D. Wilson,et al.  A smoothed EM approach to indirect estimation problems, with particular reference to stereology and emission tomography , 1990 .

[4]  R. Anderssen,et al.  A hyperplane approach to the EMS algorithm , 1992 .

[5]  Y. Vardi,et al.  From image deblurring to optimal investments : maximum likelihood solutions for positive linear inverse problems , 1993 .

[6]  H. Schneider.  Note on the Fundamental Theorem on Irreducible Non-Negative Matrices , 1958, Proceedings of the Edinburgh Mathematical Society.

[7]  Henryk Minc,et al.  On the Maximal Eigenvector of a Positive Matrix , 1970 .

[8]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[9]  L. Shepp,et al.  Maximum Likelihood Reconstruction for Emission Tomography , 1983, IEEE Transactions on Medical Imaging.

[10]  Robert S. Anderssen,et al.  On the stabilization inherent in the EMS algorithm , 1994 .

[11]  D. Nychka.  Some properties of adding a smoothing step to the EM algorithm , 1990 .

[12]  L. Shepp,et al.  A Statistical Model for Positron Emission Tomography , 1985 .

[13]  A. Brauer.  The theorems of Ledermann and Ostrowski on positive matrices , 1957 .

[14]  A. Ostrowski.  On the Eigenvector belonging to the Maximal Root of a Non-negative Matrix , 1960 .

[15]  Charles L. Byrne,et al.  Iterative image reconstruction algorithms based on cross-entropy minimization , 1993, IEEE Trans. Image Process..

[16]  P. Green.  Bayesian reconstructions from emission tomography data using a modified EM algorithm. , 1990, IEEE transactions on medical imaging.