Basics of PET and SPECT Imaging - Correction Techniques in Emission Tomography - page 61

FIGURE 4.8: Blood time activity curve for an F 18 -FDG mouse study, mea-
sured in a volume placed in the left ventricle of the mouse heart (dark gray
line). For comparison, the blood activity concentration measured in arterial
blood samples is also given (light gray line). Due to spill in from the my-
ocardium the curve measured in the images is still increasing over the time
while the real blood activity concentration decreases (left). Corresponding
FDG-PET image of a mouse with the VOI (dark gray) in the lumen of the
left ventricle where the blood activity concentration was measured (right).

can lead to wrong high values in the time activity curves as shown in Figure
4.8.
A measure for partial volume effects is the recovery coecient which is
dened as:

Measured Activity Concentration
Real Activity Concentration

RC = : (4.5)

The recovery coecient is dependent on the size of a structure, the ac-
tivity concentration inside the structure and the surrounding area and even
on the geometrical shape of the object. The size dependence of recovery co-
ecients of spheres filled with an activity concentration 10 times higher than
the surrounding background is shown in Figure 4.7 (right side). The recov-
ery coecient is dependent not only on the structure of interest but also on
the reconstruction algorithm of the images and potential post-reconstruction
filtering.

4.1.7 Time resolution and randoms (PET only)
In PET, events are measured as coincidences; the two coincidence photons
that originate from one positron decay are called true coincidences or trues
(light gray event, Figure 4.9). Due to properties of the detector material and
the electronics the time resolution of a PET scanner is limited and hence the
timing window in which the electronics consider two measured photons as a
coincident event cannot be made arbitrarily small. Typical coincidence win-