Group for Research in Decision Analysis

G-2016-25

Revisiting motion compensation models in pet image reconstruction

, , and

Positron emission tomography (PET) image reconstruction in the presence of periodic motion, such as heartbeat and breathing, has been actively investigated since the introduction of gated data acquisition.
While each gated sinogram can be reconstructed independently, better images can be obtained by incorporating all gated sinograms, which are considered statistically independent, into an iterative reconstruction scheme.   
However, this gain in image quality is achieved at the expense of substantial extra computational resources to handle all gated sinograms simultaneously. 
We provide a comparative study with a model that is built upon the interdependence of the parameters defining the stochastic processes behind each gated sinogram to yield a single merged sinogram.
We show, using simulated and experimental datasets, that images of similar quality than with the classical model can be produced, while requiring much less computing resources.

, 11 pages