To analyze signals measured from human blood flow in the time-frequenc
y domain, we used the wavelet transform which gives good time resoluti
on for high-frequency components and good frequency resolution for low
-frequency components. Five characteristic frequency peaks, correspond
ing to five almost periodic rhythmic activities, were found on the tim
e scale of minutes. These oscillations were characterized by time and
spatial invariant measures. The potential of this approach in studying
the blood-flow dynamics was illustrated by revealing differences betw
een the groups of control subjects and athletes. (C) 1998 Society for
Mathematical Biology.