Biomedical Engineering Reference
In-Depth Information
such as cerebral blood flow or the density of a receptor in a certain area, MRI
defines different structures or tissue types, and thus provides information
about morphology and the topology of structures. Both imaging modalities
are useful for clinical studies as well as for basic scientific investigations.
Once the respective images from PET and MRI have been realigned and
registered, they can be fused for an integrated display, offering the opportu-
nity to delineate function and morphology at the same time. In addition,
enhanced display options can be applied to generate three-dimensionally
rendered images, like the surface of the human cortex (extracted from the MR
images) with activated sites (extracted from the PET images) superimposed.
1
The clinical applications of PET-MRI registration will be discussed in this
chapter with a focus on practical issues, such as methods and procedures for
performing image registration with PET and MRI. Related subject matter is
treated in a complementary manner in Chapter 11. The relevant question is
which techniques and protocols have been shown to be applicable in a clini-
cal setting with its specific requirements on robustness, practicability, and
patient handling. While image registration has been an essential and, hence,
accepted step in the analysis of multimodal brain images, especially for
research purposes, only a relatively small number of publications deal with
applications to other parts of the body. Limitations experienced during
attempts to register nonbrain images will be outlined, as well as suggested
procedures for harnessing the advantages of image registration to support
the decision making process in a clinic before the start of therapy. Before
images from PET and MRI are discussed in detail within the framework of
image registration, the most important properties of the images from both
modalities are summarized.
9.2
Properties of PET Images
Positron emission tomography employs the main features of tracer tech-
niques developed to study the underlying mechanisms of physiological and
biochemical processes in a living organism (see also Chapter 11). Labeling is
obtained by exchanging one of the tracer molecule's atoms by its radioactive
analogue. The radioactively labeled substance is injected intravenously and
can be traced through the body using external detectors. In the case of PET,
the tracer is labeled with an isotope that emits a positron. Such isotopes are
available for a number of biologically relevant atoms, namely oxygen (as
O-15, t
=
2.05 min), carbon (as C-11, t
=
20.4 min), and nitrogen (as N-13,
1
2
1
2
t
109.7 min) can be used to replace
an OH-group in a molecule. Labeling with a radioactive nuclide allows the syn-
thesis of specific tracers, which are used to determine, for example, cerebral
blood flow or glucose consumption in the human brain. Hence, PET is repre-
sentative of a functional imaging modality, and the primary interest in the
development of PET was to quantify the three-dimensional distribution of
=
10 min). In addition, fluorine (F-18, t
=
1
2
1
2
