Output details

Fast generation of 4D PET-MR data from real dynamic MR acquisitions

We describe a novel method for accurately and rapidly simulating dynamic PET datasets. Basing the simulation on real dynamic MRI data allows the wide diversity in anatomy and motion patterns seen in real patients to be accounted for. Analytic simulation (100-1000 times faster than standard Monte-Carlo approaches) allows production of the very large numbers of noise realizations required for rigorous statistical assessment of PET reconstruction and motion correction algorithms. The method is implemented using the open-source ‘Software for Tomographic Image Reconstruction’ package and dynamic MR datasets with associated PET simulations have been made freely available (10 downloads requested to date).