Introduction - Correction Techniques in Emission Tomography - page 2

intra-subject comparison only after application of these corrections. For PET
imaging, for instance, this includes corrections for radioactive decay, random
and scattered events, corrections for partial volume effects, compensation for
patient, heart and breathing motion, attenuation correction, dead time cor-
rection, etc.
Additionally, hybrid emission tomography (e.g., PET-CT, PET-MRI,
SPECT-CT, PET-OI), which is increasingly being used due to the opportu-
nities presented by hybrid imaging, necessitates another range of corrections,
such as attenuation correction of PET data based on CT or MRI information.
This topic deals with the required correction methods from the point of
view of computer science, mathematics and physics. The objective is to present
an overview of relevant problems in emission tomography, possible artifacts
caused by them, and correction techniques to improve the resulting images.
1.2 Principle of emission tomography
Emission tomography is an imaging technique where a three-dimensional
distribution of photons is emitted by sources such as radionuclides or fuores-
cent compounds that are injected into the biological or human object. These
sources are part of the electromagnetic spectrum covering a large variety of
waves with different wavelengths and properties.
In emission tomography the radiation sources are transferred into an object
of interest (e.g., the human body or a mouse) by different procedures, in
most cases by injection into the blood stream. After distribution, the injected
substance (probe) emits electromagnetic waves which can then be detected
from outside the object by dedicated detectors.
Unlike X-ray imaging where the images show the absorption of X-rays
traveling through an object, emission imaging makes use of emitting probes
that have the capability of penetrating tissue. Whereas in X-ray imaging con-
trast agents are often utilized to enhance the images, the emitting probes
themselves build the contrast in emission imaging.
The most interesting but often most challenging task is the development of
specific probes where the radiation-emitting substance is chemically bound (la-
beled) to a molecule of interest. After injection of this labeled compound into a
biological system, one can follow (trace ) noninvasively the labeled molecule of
interest. This phenomenon is called the tracer principle which was first discov-
ered by George Charles de Hevesy. Due to this behavior, emission tomography
is often referred to as a molecular imaging technique. Consequently, emission
images show molecular or functional information rather than morphological
or structural details. This has led to the development of hybrid imaging tech-
niques, such as PET-CT or SPECT-CT, which combine high-level molecular