PARAMETRIC IMAGE-RECONSTRUCTION USING SPECTRAL-ANALYSIS OF PET PROJECTION DATA

Spectral analysis is a general modelling approach that enables calcula tion of parametric images from reconstructed tracer kinetic data indep endent of an assumed compartmental structure. We investigated the vali dity of applying spectral analysis directly to projection data motivat ed by the advantages that: (i) the number of reconstructions is reduce d by an order of magnitude and (ii) iterative reconstruction becomes p ractical which may improve signal-to-noise ratio (SNR). A dynamic soft ware phantom with typical 2[C-11]thymidine kinetics was used to compar e projection-based and image-based methods and to assess bias-variance trade-offs using iterative expectation maximization (EM) reconstructi on. We found that the two approaches are not exactly equivalent due to properties of the non-negative least-squares algorithm. However, the differences are small (<5%) and mainly affect parameters related to ea rly and late time points on the impulse response function (K-1 and, to a lesser extent, VD). The optimal number of EM iterations was 15-30 w ith up to a two-fold improvement in SNR over filtered back projection. We conclude that projection-based spectral analysis with EM reconstru ction yields accurate parametric images with high SNR and has potentia l application to a wide range of positron emission tomography ligands.