Partial Volume Correction of the Image Derived Input Function for Radiotracers Which Adhere to Blood Vessel Walls: Application to [¹¹C]PK11195.

Objectives: An Image derived Input Function (IDIF) can be used for the quantification of physiological processes with Positron Emission Tomography (PET) when obtaining an arterial input function is not practical due to the risk, time and cost associated with the procedure. However the IDIF may not be accurate if there is uptake of the radiotracer into the blood vessel walls as is the case for [¹¹C]PK11195. We used a simple model to correct for uptake in vessel walls to provide a more accurate IDIF.

Methods: Twenty healthy volunteers got 60 minute PET scans after the IV administration of [¹¹C]PK11195 and arterial samples were acquired at frequent intervals during the scans. The partial volume correction was modeled by comparing the arterial input function with the IDIF using the equation. A(t) = RC·Cp(t) + SF_b∙B(t) + SF_v∙(k₁Cp(t) - k₂W(t)) where A(t) is the measured blood pool concentration, Cp(t) is the true concentration, B(t) is the surrounding tissues activity and W(t) in the vessel wall. SF_b is the spill in factor from the background, SF_v is the spill in factor from the vessel walls. The terms k₁ and k₂ are the influx and efflux rates of the radiotracer with the vessel wall. RESULTS: The findings showed that an added spill-in factor was required to get good agreement between the AIF and the IDIF. This factor was most likely due to the build-up of the radiotracer in the walls of the blood vessels and was modeled using first order kinetics driven by the concentration of the radiotracer in the blood pool. In a statistical comparison of the input functions in the 20 subjects using the diameter of the carotid artery as the only variable, the mean value a paired two-tailed T test gave a p value of 0.34 indicating that there was no statistically significant difference between the AIF and the fully corrected IDIF. CONCLUSION: The accumulation of radiotracer in the walls of the vessel can play an important part in accurately determining the IDIF. A separate factor can be introduced in the standard model to account for this spill-in. This could improve bias and variance in studies that use an IDIF instead of an AIF for quantification.