Image segmentation and activity estimation for microPET 11C-raclopride images using an expectation-maximum algorithm with a mixture of Poisson distributions

Abstract

The objective of this study was to use a mixture of Poisson (MOP) model expectation maximum (EM) algorithm for segmenting microPET images. Simulated rat phantoms with partial volume effect and different noise levels were generated to evaluate the performance of the method. The partial volume correction was performed using an EM deblurring method before the segmentation. The EM-MOP outperforms the EM-MOP in terms of the estimated spatial accuracy, quantitative accuracy, robustness and computing efficiency. To conclude, the proposed EM-MOP method is a reliable and accurate approach for estimating uptake levels and spatial distributions across target tissues in microPET (11)C-raclopride imaging studies.