Abstract

A model for quantification of perfusion in pulsed labelling techniques is described, based on solving the modified Bloch equation including the effects of flow. The model is designed to fit experimental data acquired in two separate measurements (inversion and control, or selective and non-selective inversions) for different inversion times using a biexponential. Although the signal contrast is 50% less than the continuous labelling technique, it seems more appropriate for human studies because of its lower power deposition, shorter transit time and the use of an interleaved acquisition. The importance is shown of including in the model the difference in relaxation time between blood and tissue. Neglecting this difference can lead to an overestimation of flow, which can be as big as 100% in white matter and 20% in grey matter.