Basics of PET and SPECT Imaging - Correction Techniques in Emission Tomography - page 54

TABLE 4.1: Linear photon attenuation coecients at 140 keV and 511 keV
for different materials (from ICRU Report 44 [1] and Hubbell 1969 [3]).

Density at 140 keV at 511 keV

[g/cm 3 ]

Material [1/cm] [1/cm]

Water 1.00 0.150 0.095

Lung 1.05 0.04-0.06 0.025-0.04

Fat tissue 0.95 0.142 0.090

Cortical bone 1.92 0.284 0.178

Muscle 1.05 0.155 0.101

ent materials (Z independent) because electron density is similar for most
materials.

4.1.2 Photon attenuation
Photons, in the energy range of SPECT as well as for the 511 keV coinci-
dence photons measured in PET, interact within the patient either via photo-
electric effect or Compton scattering. In the case of absorption, the photons
will not be detected and their information is lost for the image reconstruction.
In the case of scattering, the photon may still be measured in the detector,
leading to wrong information for the image reconstruction. This is addressed
in the next paragraph (4.1.3). All processes that reduce the number of de-
tected photons are put together under the term attenuation. The degree of
attenuation of a photon beam depends on the material through which the
beam traverses and on the thickness of the material. The attenuation capac-
ity of a material is described by its linear attenuation coecient . Values for
some materials can be found in Table 4.1.
Attenuation for a photon beam starting at s 0 can be described with the
following formula:

TF(s 0 ) = e R s 0 (s)ds :

(4.3)

While for small animal imaging photon attenuation is often neglected as
the effect is quite small as the photons are attenuated by only a few millimeters
of tissue, for human PET and SPECT, photon attenuation in the patient
leads to a massive underestimation of the activity concentration in PET data,
interfering with quantitative and even qualitative image analysis, as discussed
later. In Figure 4.2 (top and bottom) effects of photon attenuation in PET
can be seen in a cylindrical phantom (20 cm diameter) that was filled with a
homogeneous activity concentration. In the center of the phantom the activity
concentration seems to be lower by a factor of 10 to 11, compared to the
image corrected for attenuation. Besides the false absolute quantification of
the radiotracer uptake, the different attenuation in the center and the outer
parts of the body can lead to problems in the visual image, too. In patients,