Cerebral blood flow studies in man using recirculation corrected height/area (H/A) computations for intravenous injection of 133Xenon.

Our height/area (recirculation-corrected) non-invasive cerebral blood flow method compares well with k1 values and k average as determined from k1, k2, p1 and p2 by Obrist computations. The height/area method provides regional flow measurements with less variability than k1 due to low regional head counts, head motion and end-tidal gas sample variations. Derivation of the mathematical model is for a single flow compartment and suited for tomographic measurements such as PET. Collimated detector data using our computation provides a good approximation of flow even for multiple compartments. Our approach permits a lower counting rate by either a smaller isotope dose or better detector collimation than multicompartment curve fitting methods.