Appropriate Methodology for Automated Scaling of DSC-CBF Images for Stroke Evaluation

Jeffry R. Alger^1,2, T. J. Schaewe, J. J. Wang, D. S. Liebeskind, Q. Hao, J. X. Qian², J. L. Saver, N. Salamon², UCLA Stroke Investigators

The study evaluated automatic approaches to selection of arterial input function (AIF) and the value of AIF correction with a measured venous outflow function (VOF). Automated AIF selection using angiographic criteria was moderately more optimal than AIF selection based on dynamic criteria alone. VOF correction was helpful for both types of AIF sampling. Even with these optimal procedures, a variance of more than 40% in the derived most probable CBF measured in normal brain can be anticipated in routine clinically-indicated DSC studies.