Andrea Kassner1,2, Rebecca E. Thornhill1,2,
Swati Matta1, Fang Liu1, David J. Mikulis1,3
1Medical Imaging, University of
Toronto, Toronto, Ontario, Canada; 2Physiology and Experimental
Medicine, Hospital for Sick Children, Toronto, Ontario, Canada; 3Medical
Imaging, Toronto Western Hospital, Toronto, Ontario, Canada
Thrombolytic
therapy is known to increase the risk of hemorrhagic transformation (HT) in
acute ischemic stroke (AIS). Accurate
and robust methods for predicting HT are required for improving treatment
guidance. Model-based permeability
estimation with dynamic contrast-enhanced MRI can predict HT, but the
estimates (KPS coefficients) are sensitive to noise and require an arterial
input function. However, studies of tumors suggest that a model-free measure,
the initial area under the contrast-concentration curve (IAUC) is more
robust. We evaluated both KPS and IAUC in AIS patients and found that only
KPS successfully delineated HT. Model-based estimates are recommended over
IAUC in AIS.