Correcting for the echo-time effect after measuring the cerebral blood flow by arterial spin labeling

Abstract

Purpose: To take into account the echo time (TE) influence on arterial spin labeling (ASL) signal when converting it in regional cerebral blood flow (rCBF). Gray matter ASL signal decrease with increasing TE as a consequence of the difference in the apparent transverse relaxation rates between labeled water in capillaries and nonlabeled water in the tissue (δR2*). We aimed to measure ASL/rCBF changes in different parts of the brain and correct them. Materials and Methods: Fifteen participants underwent ASL measurements at TEs of 9.7-30 ms. Decreases in ASL values were localized by statistical parametric mapping. The corrections assessed were a subject-per-subject adjustment, an average δR2* value adjustment, and a two-compartment model adjustment. Results: rCBF decreases associated with increasing TEs were found for gray matter and were corrected using an average δR2* value of 20 s -1. Conversely, for white matter, rCBF values increased with increasing TEs (δR2* = -23 s -1). Conclusion: Our correction was as good as using a two-compartment model. However, it must be done separately for the gray and white matter rCBF values because the capillary R2* values are, respectively, larger and smaller than those of surrounding tissues. Copyright © 2011 Wiley-Liss, Inc.