[A fast technique to estimate local cerebral blood flow and partition coefficient using dynamic PET of H(2)15O: a new approach to weighted integration method based on time integration of Kety-Schmidt equation].

We developed a new technique for fast estimation of local cerebral blood flow and partition coefficient based on time integration of Kety-Schmidt equation. This technique has two advantages concerning statistical noise. Firstly, it does not need a Ci(T) image which contains high statistical noise because of low activity in brain. Secondly, statistical noise is minimized by time integration of Kety-Schmidt equation before multiplied by weighting function. The error analyses in flow and partition coefficient caused by statistical noise were performed using computer simulation, and dynamic PET human study using H(2)15O bolus injection. Both simulation and PET study indicated that uncertainties of flow and partition coefficient were confirmed to be lower in this technique than in the original weighted integration method.