Birgitte Fuglsang Kjolby1, Irene Klrke Mikkelsen1, Michael Pedersen2, Leif Oestergaard1, Valerij G. Kiselev3
1Center for Functionally Integrative Neuroscience, Aarhus University Hospital, Aarhus, Denmark; 2MR Research Center, Aarhus University Hospital, Aarhus, Denmark; 3Department of Diagnostic Radiology, Medical Physics, University Hospital Freiburg, Freiburg, Germany
Quantitative perfusion weighted MRI requires an accurately measured arterial input function (AIF). Due to limited spatial resolution, AIF voxels cannot be placed completely within a feeding artery. We present a two-compartment model of an AIF voxel including relaxation properties of blood and tissue. The aim is to analyze the shape of partial volumed AIF signals and quantify the impact of partial volume effects on the quantitative perfusion metrics. We find that the tissue contribution broadens and introduces fluctuations in the AIF. Furthermore, PVE bias perfusion metrics in a nonlinear fashion, compromising quantitative perfusion estimates and profoundly effecting local AIF selection.