The requirement for constant arterial radioactivity in the C15O2 steady-state blood-flow model.

This study evaluates the discrepancy between the true CBF value and the CBF value calculated according to the C15O2 steady-state model, for situations where the arterial input function, Ca(t), deviates considerably from its steady-state value, Ca. The fact that arterial input function and tissue O-15 concentration are not independent variables is taken into account. Inconstant or variable arterial input functions are simulated and the corresponding tissue O-15 concentrations calculated. The steady-state CBF values are evaluated for several temporal variations of Ca over the period of imaging, all derived from Ca(t) by simulation of various blood-sampling schemes, and are compared with the true CBF value. The study indicates that reliable CBF values are obtained by the C15O2 steady-state method even under severely impaired "unsteady-state" conditions, provided that either the true average arterial concentration over the entire scan, or the average concentration from multiple arterial samples, is used.