Biomedical Engineering Reference
In-Depth Information
therefore to transform accurately the CT image data acquired at an effective
energy to proper attenuation coecient values at 511 keV. Unfortunately,
knowledge of
CT
alone is not sucient to determine the proper value of
PET
. This is basically because the contribution of Compton scattering to the
attenuation coecient is depending on the (electron) density of the material
alone, while the contribution of the photoelectric effect is additionally strongly
dependent on the effective atomic mass Z
eff
of the material. Therefore, ma-
terials having the same
CT
may actually have different
PET
[86]. However,
the composition of human bodies usually allows additional constraints leading
to a basis for a reliable transformation, as in terms of different attenuation
behavior, the human body basically consists of three materials and mixtures
thereof: air, water, and bones. Soft tissue with varying density can be well
modeled as an accordingly varying mixture of water and air. This simple de-
composition into mixtures of air/water and water/bone compartments leads
to the following scaling between CT and PET attenuation values
HU + 1000
1000
PET
(HU 0) =
PET
H
2
O
·
(5.18)
HU
PET
Bone
PET
H
2
O
PET
(HU > 0) =
PET
H
2
O
+
PET
1000
CT
Bone
C
H
2
O
·
(5.19)
H
2
O
with predetermined values for
PET
H
2
O
,
PET
Bone
,
C
H
2
O
, and
CT
Bone
(see Figure 5.11)
[15]. It should be mentioned that a change in effective X-ray energy (by chang-
ing tube voltage settings) requires an adjustment of these values. Similar trans-
formations have been introduced for scaling CT values to SPECT attenuation
values [12] [83]. Alternatives to bilinear transformations have been suggested
[43].
Although now often considered a gold standard in clinical practice, CT-
based attenuation correction approaches in emission tomography potentially
introduce new sources of image artifacts and erroneous tracer quantification.
As these problems are important subjects of current research, they are be
discussed in detail in the next section.
5.4.3 Attenuation correction artifacts
As already mentioned, the main advantages of CT-based attenuation cor-
rection over transmission scan-based attenuation correction using external ra-
dionuclide sources are time eciency (a CT scan can be performed within sec-
onds, while an ordinary transmission scan usually takes at least a few minutes)
and low noise properties of the acquired image data due to the much higher
photon flux during the CT scan as compared to conventional radionuclide
transmission scans. The latter point also allows post-injection transmission
scans in PET/CT and SPECT/CT systems without correcting for emission
event contamination which can be a problem in radionuclide-based transmis-
sion scans as described above.
Search WWH ::
Custom Search