Biomedical Engineering Reference
In-Depth Information
2.11 Data Corrections
Since one of the unique features of PET is its ability to provide quantitative
images that are directly related to the physiology of the process under study,
accurate data acquisition and corrections are required before or during the re-
construction process in order to achieve absolute or relative quantification.
2.11.1 Detector Normalization
A modern PET scanner consists of multiple rings of many thousands of detector.
It is not possible that all detectors have the same operation characteristics due
to differences in exact dimensions of the detectors, the optical coupling to the
PMTs, and the physical and geometrical arrangement of the detectors. In other
words, it means that different detector pairs in coincidence will register different
counts when viewing the same emitting source. Therefore, the entire set of
projection data must be normalized for differences in detector response. The
normalization factors can be generated for each coincidence pair by acquiring
a scan in the same way as blank scan, with a rotating rod source of activity
orbits at the edge of the FOV of the gantry. Adequate counts must be acquired to
prevent noise propagation from the normalization scan into the reconstructed
image.
2.11.2 Dead-Time Correction
During the period when a detector is processing the scintillation light from a
detected event, it is effectively “dead” because it is unable to process another
event. Since radioactive decay is a random process, there is a finite probabil-
ity that an event occurs at a given time interval. If an event occurs during the
interval when the detector is “dead,” it will be unprocessed, resulting in a loss
of data. Such loss of data is referred to as dead-time loss . As count rate in-
creases, the probability of losing data due to dead-time increases. Dead-time
losses are not only related to the count rates but also depend upon the analog
and digital electronic devices of the system. To correct for dead-time, one can
plot the measured count rate of a decaying source over time. If the source is
a single radionuclide, one can calculate the count rate from the half-life of the
Search WWH ::




Custom Search