Abstract #0376

Subject-specific AIF Optimizes Reproducibility of Perfusion Parameters in Longitudinal DSC-MRI in Comparison to Session & Population Level AIF

Kim Mouridsen¹, Kyrre Eeg Emblem², Atle Bjrnerud³, Dominique Jennings², Gregory Sorensen²

¹Center for Functionally Integrative Neuroscience, Aarhus University | Aarhus University Hospital, Aarhus, Denmark; ²Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Charlestown, MA, USA; ³Department of Physics, University of Oslo, Interventional Centre, Rikshospitalet, Olso University Hospital, Oslo, Norway

DSC-MRI is used intensively in imaging studies tracking the effect of emerging treatment paradigms, e.g. anti-VEGF therapy in tumors. It is crucial to the success of longitudinal studies to minimize intrapatient reproducibility of perfusion indices, such as CBF, CBV and MTT. We demonstrate substantial improvements in reproducibility in all parameters by revising the AIF search strategy, estimating a patient specific AIF rather than determining the AIF at each scan. We also observe a lower reproducibility using a population-based AIF. This suggests that a patient-level AIF correctly adjusts for variations in systemic circulation while minimizing scan-rescan variability in perfusion values.