Abstract #0795

Parker G, Buonaccorsi G

Analytical techniques for DCE-MRI would benefit from validation against ground truth, but patient data is generally unsuitable for this purpose. We have therefore developed a framework for generating synthetic DCE-MRI data sets that includes the incorporation of an appropriate kinetic model and the addition of motion and noise. We have applied tracer kinetic model fitting and kinetic model-driven registration to a simple synthetic data set generated using this framework, with positive results. We propose that synthetic data of this type will prove invaluable for validating DCE-MRI techniques, particularly when these are to be used in clinical trials.