Joint estimation of the activity and the events of interaction in SPECT systems

We describe a unified probabilistic graphical model for joint reconstruction of the activity and of the coordinates of the events of interaction in Emission Imaging. This new formulation is a discrete-discrete alternative to the discretecontinuous List Mode Likelihood model described by Barrett et at. The discrete-discrete formulation allows us to account for nonhomogeneous resolution of the gamma-camera while adopting computationally efficient shift-invariant projection and backprojection algorithms. We describe an application of the model for the compensation of disrupted photo-detectors is SPECT.

[1]  Stefano Pedemonte,et al.  Optimum real-time reconstruction of Gamma events for high resolution Anger camera with the use of GPGPU , 2009, 2009 IEEE Nuclear Science Symposium Conference Record (NSS/MIC).

[2]  Radford M. Neal.  Pattern Recognition and Machine Learning , 2007, Technometrics.

[3]  L.R. Furenlid,et al.  Real-time data acquisition and maximum-likelihood estimation for gamma cameras , 2005, 14th IEEE-NPSS Real Time Conference, 2005..

[4]  H. Barrett,et al.  List-mode likelihood: EM algorithm and image quality estimation demonstrated on 2-D PET , 1998, IEEE Transactions on Medical Imaging.

[5]  H H Barrett,et al.  List-mode likelihood. , 1997, Journal of the Optical Society of America. A, Optics, image science, and vision.

[6]  J. Besag.  On the Statistical Analysis of Dirty Pictures , 1986 .

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