A new method is presented for tracking the motion of blood and determining
its velocity spectrum from magnetic resonance data collected within a singl
e heartbeat. The method begins by tagging a column of blood in a vessel by
combining a 1D SPAMM excitation with a 2D cylindrical excitation. A series
of 1D projections of the tagging pattern is acquired from a train of gradie
nt echoes. The influence of specific excitation profiles and velocity profi
les on the motion of the tags is explored for steady flow. It is shown math
ematically, and confirmed with phantom experiments, that the velocity of a
tag equals the mean velocity of the excited fluid when the velocity spectru
m is symmetric about its mean velocity. The velocity spectrum is derived by
analyzing the interference between tags moving at different velocities. Th
is appears to be the first use of magnitude tagging to obtain velocity spec
tra. Representative measurements in a human aorta are presented to assess f
easibility in vivo. (C) 2001 Wiley-Liss, Inc.