Comparison of sinogram and reconstruction based correction for partial volume effects in ECT imaging

The accurate quantification of single photon emission computed tomography (SPECT) and positron emission tomography (PET) data is limited by partial volume effects (PVE). This study compares two approaches to measure accurately the true tissue tracer activity within a tissue compartment, defined by anatomical side information. The first approach is based on linear regression analysis, performed on rescaled reconstructed ECT images. The second method is based on linear regression analysis of the projection data. Both methods can also estimate the activity of a tissue compartment locally by linear regression analysis taking into account only the nearest neighbors. A simple simulation study shows that the techniques based on reconstructed intensity values yields a more accurate quantification for a small hot compartment, surrounded by both a background compartment with a homogenous activity distribution. Applied locally, a more accurate quantification is achieved in the case of a background compartment with an inhomogeneous activity distribution. Key-Words: SPECT, PET, quantification, partial volume effects, linear regression.