Advanced-Model 3D DCE-MRI of Small Animals with Spatially Regularized Processing Steps

In DCE-MRI, advanced pharmacokinetic models, such as the 2CX, TH and ATH models, provide a more complete set of pharmacokinetic parameters than the commonly used simpler models, such as the Tofts and Patlak models. However, their use requires higher temporal resolution and signal-to-noise ratio. Hence, advanced-model DCE-MRI allows for imaging of only a few slices. We present an approach making 3D advanced-model DCE-MRI possible and evaluate it under the low-signal-to-noise-ratio conditions of small-animal MRI. Our methodology is based on compressed sensing and spatially regularized fitting of the pharmacokinetic model. The approach is evaluated on simulated and real datasets.