ABSTRACT
I. Introduction 136
II. Technical Principals and Overview of Micro-CT 137
A. Micro-CT Introduction 138
B. Micro-Focal Sources of X-ray Production 140
C. Filters 140
D. Some Detector Characteristics 143
E. Comments on Dose Exposure 145
F. Resolution of the Systems 146
G. Tomographic Reconstruction 147
H. Ring Artifact Reduction 148
I. Beam Hardening 149
J. Hounsfield Unit Calibration 151
K. Commercially Available Systems 151
III. Small Animal Lung Imaging 152
A. Optimized In Situ and In Vivo Mouse Lung Protocol 154
In Situ Scanning Protocols 154
In Vivo Scanning Protocols 156
B. Examination of Lung Tissue Samples from Larger Animals 156
C. Dynamic Imaging: Microfluoroscopy Coupled to Micro-CT 158
D. Conclusions 163
References 166
I. Introduction
Computed tomography (CT) of the lung has made major advances in hardware and
software over the last decade. The human lung can now be imaged rapidly and
volumetrically, with digital information easily transported, stored, reviewed, and
then both subjectively and objectively analyzed (1,2). Figure 6.1 depicts examples
that span anatomic segmentation (3-8), automatic labeling (9), image based lung
modeling (10,11), ventilation (12,13) and perfusion imaging (14,15), a morpho-
metric correlation to in vivo imaging (16), and a large animal model of evolving
lung inflammation. In large animal models, functional imaging demonstrates
early changes in pathology (17). Our methods demonstrate that measures of both
structure and function provide early signs of pathologic processes (2). Still, there
remains a need to correlate pathologic phenotypes with genotypes and to pheno-
type diversity at the alveolar/bronchiolar and arteriolar levels in vivo and in situ. It is our view that micro-CT will have a profound influence on the docu-
mentation of anatomical and physiological phenotypic changes in many
genetic mouse models. The mapping of the human genome, together with that
of other animals and plants, is providing an enormous amount of new infor-
mation, the full extent of which will still emerge over the next several
decades. The mouse has become the prototypic animal model for the study of
genetic based diseases. Frequently, however, abnormalities are noted within
the animals, beyond the specifically induced biological defect or outside the tar-
geted organ. Thus, the full phenotype of these animals may include a variety of
unexpected anatomical and physiological abnormalities, in addition to biochemi-
cal and genetic changes. Describing these abnormalities fully is critical to under-
standing the complex interaction of different genetic influences. Although
individual animals are relatively inexpensive, the cost of developing a model
can be very significant. This expense is amplified when animals must be sacri-
ficed and studied at multiple time points during growth or during the development
of a disease. Determining such time points is somewhat arbitrary, and important
information may not be recorded if complete observations are not made. Micro-
CT scanning will allow numerous experiments in mouse models to be planned
with greater precision. The cost of exploring the complex phenotypic expression
of genetic changes will be reduced, and longitudinal studies will be greatly facili-
tated by allowing a more complete and accurate description of events. As shown
in Fig. 6.2, mouse imaging using conventional CT scanners lacks the spatial
resolution needed. To maximize the potential of micro-CT to quantitatively
evaluate the mouse lung, imaging protocols, reconstruction algorithms, and
image analysis methods all must be established and specifically tailored to
in vivo, in situ, and ex vivo imaging of lung tissue.
