The direct calculation of parametric images from raw PET data using maximum likelihood iterative reconstruction

Abstract

The aim of this work is to calculate, directly from raw projection data, clear and concise images characterising the spatial and temporal distribution of labelled compounds. Conventionally, image reconstruction and the calculation of parametric images are performed separately. By combining both processes, low noise parametric images are obtained, bypassing the expensive computations of image reconstruction. The calculation of the images is performed by restricting the pixel time activity curves to a positive linear sum of predefined time characteristics. The weights in this sum are calculated directly from the PET projection data, using an iterative algorithm derived from a Maximum Likelihood (ML) iterative algorithm commonly used for tomographic reconstruction. Combining reconstruction and parametric estimation also produces a computational advantage. The ability of the algorithm to extract known kinetic components from the raw data is assessed, using data from both a phantom experiment and from clinical studies. The calculated parametric images indicate differential kinetic behaviour and have been used to aid in the identification of tissues which exhibit variations in the handling of labelled compounds.