Page 357 - Biomedical Engineering and Design Handbook Volume 2, Applications

NUCLEAR MEDICINE IMAGING INSTRUMENTATION 335

the random coincidence rate can exceed the true coincidence rate. The random coincidences provide
false information and need to be removed from the acquired data prior to image reconstruction. It is
also obvious that random coincidence rate can be reduced with a smaller coincidence time window.
That requires detectors with a fast response time like LSO, LYOS, and GSO.
For sources in air, it is only possible to get a true coincidence event when the source lies in the
defining volume between the two coincidence detectors. However, if the sources are distributed in
some material, like human tissue, it is possible for one or both of the annihilation photons to be scat-
tered into detectors that don’t encompass the LOR of the source (Fig. 11.12b). Like the random coin-
cidence event, this provides false information that requires correction. The number of scattered
events can be reduced by energy discrimination, but this does not eliminate it all and additional scatter
correction techniques are required for PET imaging.

11.4.1 PET Scanner Design
PET imaging systems are based on coincidence detection, and it is possible to use scintillation camera
technology to create PET tomographs and for awhile, SPECT/PET hybrid systems were widely
available. The low intrinsic efficiency of the gamma cameras made these systems inferior to dedi-
cated PET systems and reimbursement for clinical imaging studies performed with SPECT/PET sys-
tems was discontinued, sealing its doom as a commercial product. One dedicated PET system that
uses the gamma camera approach is still available. In this design, there are six individual scintilla-
tion cameras with 25-mm-thick NaI(Tl) detectors. The detectors are curved so that when they are
combined, they form a complete cylinder with a diameter of 90 cm and an axial length of 25 cm. The
C-PET operates as a three-dimensional PET system and is less expensive than the ring systems with
individual detectors. It has spatial resolution that is comparable to the ring detector systems and
much better energy resolution. However, its count sensitivity is about half that of the ring systems.
The best performing whole body PET systems have a large array of scintillation detectors that
27
form a cylinder that the patient passes through. If a single ring of this system is considered, each
detector within the ring can form a coincidence event with any of the opposing detectors as shown
in Fig. 11.13. The associated lines of response provide all the necessary sampling to generate a tomo-
graphic image for one plane. To increase the number of planes, additional rings are combined
together to form a large cylinder covering 16 to 18 cm. The number of detectors used in commercial
whole body PET systems ranges from 9000 to greater than 20,000. Table 11.6 has a summary of PET
system performance specifications.
28
Four scintillators are currently being used for whole body PET systems : bismuth germanate
(BGO), lutetium oxyorthosilicate (LSO), yttrium-doped lutetium oxyorthosilicate (LYSO), and gadolin-
ium oxyorthosilicate (GSO) (Table 11.5). BGO has the best intrinsic efficiency, but its light output is
very low and the scintillation decay time is quite long. As would be expected, this results in poor energy
resolution and diminished count rate capability. The coincidence intrinsic efficiency of LSO and LYSO
is 7 percent below that of BGO, but they have very high light output and the scintillation decay time is
almost a factor of 10 shorter. One problem LSO and LYSO have that the other detectors do not share
is that they are radioactive so that the detectors have a continual background detection rate. Fortunately,
this is not a significant problem in the coincidence mode. GSO intrinsic efficiency is 22 percent less
than BGO, but it also has a much higher light output and short decay time.
As stated above, whole body PET systems can have more than 20,000 individual detectors. In the
early PET scanners, each detector was coupled directly to a single photomultiplier tube (PMT). The size
of the PMT was at that time the limiting factor to detector size and therefore the limiting factor to PET
resolution. In the 1980s, the concept of the detector block was developed (Fig. 11.14). In this scheme,
an 8 × 8 array of detectors is coupled to four PMTs. Each detector in the array is accurately identified
from the ratios of the difference and sums of the PMT signals. This innovation was very important since
it provided an economical solution to reducing the detector size while preserving count sensitivity. Block
detectors are still in use by two of the commercial vendors. The other major vendor uses an approach
similar to the Anger logic used on a scintillation camera. In their design, an array of PMTs views the
detector matrix. When an event occurs, the involved detector is determined from summing the position