The factor analysis methodology for use in the quantification of dynamic H215O-PET studies is based on the work of F. Hermansen and co-workers from the Hammersmith Hospital, London, UK [3,4,5,27]. Note that for this analysis PET transmission images with exactly the same geometry as the emission images can be used.
Principle of the Factor Analysis (FA)
The principle of the cardiac H215O-PET factor analysis is model-based and consists of the following steps:
1.An estimate of the blood H215O activity concentration is derived from a lung VOI.
2.The expected uptake and washout in myocardium is calculated based on a 1-tissue compartment model, the blood activity as the input curve, and an average myocardial perfusion value.
3.The two types of time courses (blood, myocardium) are fed into a factor analysis procedure which calculates two weights per time point, the myocardial mi factors and the blood factors bi.
4.A myocardium factor image can then be obtained by summing the frames weighted by the mi factors, and a blood factor image by summing the frames weighted by the bi factors.