Biomedical Engineering Reference
In-Depth Information
achieved popularity. It has been used widely in myocardial imaging in both
SPECT (perfusion) and PET (myocardial blood flow, -DG uptake, adreno-
receptors). The method has also been applied for the comparison of PET and
SPECT cardiac data in the same subject.
54
Spatial normalization is not only used in myocardial scanning. A common
application in neuroscience research is to assess the change in a functional
parameter (cerebral perfusion, cerebral glucose metabolism) in response to
some challenge or stimulation such as a motor, cognitive, or sensory task.
These “cerebral activation” studies demonstrating functional anatomy have
been widely reported in both the scientific and popular literature and are dis-
cussed in Chapter 14. As in the case of myocardial perfusion scanning, the
data need to be spatially normalized to a standard space to permit compar-
isons with scans from other individuals. Friston et al. chose a neuro-ana-
tomical atlas of a normal brain published by Talaraich and Tournoux
55
as the
standard space for their PET scan analysis.
56
Much effort has gone into pro-
ducing high quality standard atlases for neuroimaging,
57-59
and much of this
work continues. Functional anatomical studies of the brain now employ such
diverse measurement devices as SPECT, PET, functional MRI (fMRI), ERP
(event-related potential measured using EEG), and magneto-encephalography
(MEG). Standardization of the space for reporting the data from all of these
devices is necessary to provide a common platform for discussing results in
this new, emerging field. A standard model has also been developed for the
airways of the lung,
60
which has application in the modeling of the distribu-
tion of inhaled radioactively labeled particles throughout the airways.
61
The
aim of standardization in these applications is to aid in targeting drugs to
specific generations of the airways.
18
[]
F
11.6
Conclusions and Future Directions
Coregistration with images from complementary modalities has been
employed in nuclear medicine for many decades. It acts as an adjunct to
interpretation of the functional nuclear medicine images, as well as offering
the ability to overcome some intrinsic limitations in nuclear medicine images.
We are currently witnessing an increasing convergence in the combination of
structural and functional data, most notably in the development of dual
modality imaging devices. Even with single modality devices, we will see
further developments of algorithms and software to enhance the information
provided in combination with other complementary data.
The applications and use of image registration in nuclear medicine in the
near future will include:
•
Correlative image interpretation
•
Attenuation correction
