Background: Kinetic modeling of physiological function using imaging techniques requires the accurate measurement of the time-activity curve of the tracer in plasma, known as the arterial input function (IF). The measurement of IF can be achieved through manual blood sampling, the use of small counting systems such as beta microprobes, or by derivation from PET images. Previous studies using beta microprobe systems to continuously measure IF have suffered from high background counts.
Methods: In the present study, a light-insensitive beta microprobe with a temporal resolution of up to 1 s was used in combination with a pump-driven femoral arteriovenous shunt to measure IF in rats. The shunt apparatus was designed such that the placement of the beta microprobe was highly reproducible. The probe-derived IF was compared to that obtained from manual sampling at 5-s intervals and IF derived from a left ventricle VOI in a dynamic PET image of the heart.
Results: Probe-derived IFs were very well matched to that obtained by "gold standard" manual blood sampling, but with an increased temporal resolution of up to 1 s. The area under the curve (AUC) ratio between probe- and manually derived IFs was 1.07 ± 0.05 with a coefficient of variation of 0.04. However, image-derived IFs were significantly underestimated compared to the manually sampled IFs, with an AUC ratio of 0.76 ± 0.24 with a coefficient of variation of 0.32.
Conclusions: IF derived from the beta microprobe accurately represented the IF as measured by blood sampling, was reproducible, and was more accurate than an image-derived technique. The use of the shunt removed problems of tissue-background activity, and the use of a light-tight probe with minimal gamma sensitivity refined the system. The probe/shunt apparatus can be used in both microprobe and PET studies.