schambach 2010
TRANSCRIPT
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 1/12
Application of micro-CT in small animal imaging
Sebastian J. Schambach a, Simona Bag a, Lothar Schilling b, Christoph Groden a, Marc A. Brockmann a,*
a University of Heidelberg, Medical Faculty Mannheim, Department of Neuroradiology, Germanyb University of Heidelberg, Medical Faculty Mannheim, Division of Neurosurgical Research, Germany
a r t i c l e i n f o
Article history:
Accepted 21 August 2009
Available online 23 August 2009
Keywords:
Micro-computed tomography (lCT)
Imaging
Preclinical studies
Rodent
Mouse
Rats
Mice
Small animal
In vivo studies
a b s t r a c t
Over the past decade, the number of publications using micro-computed tomography (lCT) imaging in
preclinical in vivo studies has risen exponentially. Higher spatial and temporal resolution are the keytechnical advancements that have allowed researchers to capture increasingly detailed anatomical
images of small animals and to monitor the progression of disease in small animal models. The purpose
of this review is to present the technical aspects of lCT, as well as current research applications. Our
objectives are threefold: to familiarize the reader with the basics of lCT techniques; to present the type
of experimental designs currently used; and to highlight limitations, future directions, in lCT-scanner
research applications, and experimental methods. As a first step we present different lCT setups and
components, as well as image contrast generation principles. We then present experimental approaches
in order of the evaluated organ system. Finally, we provide a short summary of some of the technical
limitations of lCT imaging and discuss potential future developments in lCT-scanner techniques and
experimental setups.
2009 Published by Elsevier Inc.
1. Introduction
Small animals are essential as models of human disease and the
study of organism development. Small animal imaging has a vital
role in understanding these models and a key role in phenotyping,
as well as drug development and treatment. In the early 1970s,
clinical imaging was revolutionized by the introduction of com-
puted tomography (CT). Until then, the examination of small
rodents in research projects, especially of mice and rats, was
limited by the relatively low geometrical resolving capacity of
clinical CT scanners to 1 mm3 [1]. Over the past three decades,
micro-CT (lCT) imaging has rapidly advanced with higher quality
resolution, the introduction of the cone beam reconstruction
algorithm, and an increased availability of dedicated scanners for
non-invasive small animal imaging research [2]. This increased
use of lCT has been reflected in a rising number of publicationsbeginning in the early 1980s. Fig. 1 graphically depicts this rising
number of annual publications of lCT in preclinical research,
underlining the increased importance of these scanners. This graph
is based on a simple query of the public database PubMed using
the mesh terms: lCT or MICRO-CT or ‘‘High Resolution CT” or
Mini-CT and ANIMAL.
Initially lCT demonstrated excellent spatial resolution, but poor
soft tissue contrast. Therefore, early publications implementing
lCT mainly focused on the non-invasive evaluation of high con-
trast structures, such as bones or implants. With advancements
in X-ray detector sensitivity, notable improvements were made
both in temporal and in geometrical resolution, as well as readout
speed. In addition, with the introduction of new contrast agents to
elevate soft tissue contrast, lCT could be transferred to in vivo
applications in preclinical research to evaluate soft tissue
structures and vessel morphology.
The primary purpose of this review is to familiarize the reader
with the underlying technical aspects and application possibilities
of lCT imaging in experimental small animal imaging. The objec-
tives of this review are threefold: first, to present the technical fun-
damentals of lCT; second, to describe successfully applied
experimental lCT setups including various contrast generationmechanisms and contrast enhancement possibilities in relation
to the examined organ system; and finally, to identify current
limitations of lCT-imaging and future directions.
2. Technical aspects of lCT imaging
Since the first description of lCT use in preclinical research in
the early 1980s [3–6], a number of reviews of the technology and
applications of lCT have been published [1,7–15]. Initially, numer-
ous small companies specialized in the production of dedicated
small animal lCT scanners, but they were subsequently bought
by larger competitors with growing interests in lCT technology.
1046-2023/$ - see front matter 2009 Published by Elsevier Inc.doi:10.1016/j.ymeth.2009.08.007
* Corresponding author. Address: University of Heidelberg, Medical Faculty
Mannheim, Department of Neuroradiology, Theodor-Kutzer-Ufer 1-3, 68167
Mannheim, Germany. Fax: +49 621 383 2165.
E-mail address: [email protected] (M.A. Brockmann).
Methods 50 (2010) 2–13
Contents lists available at ScienceDirect
Methods
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y m e t h
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 2/12
General Electric acquired Enhanced Vision System Corp. (EVS)
in the year 2002; Siemens bought ImTek Inc. in 2004 and CTI
Molecular Imaging Inc. in 2005, while Varian Security & Inspection
Products acquired Bio-Imaging Research Inc. (BIR) in 2007. A short
market overview can be found in Table 1. To the best of our knowl-
edge, this is a comprehensive overview of the market at the time of
this review. We recognize, however, that other suppliers may also
exist of which we have not listed here.
Fig. 1. The rising number of annual publications of lCT in preclinical research demonstrates the increasing importance of these scanners. This graph is based on a simple
query of the public database PubMed with using the following mesh terms: lCT or MICRO-CT or ‘‘High Resolution CT” or Mini-CT and ANIMAL. A time line was created with
MEDSUM: an online MEDLINE summary tool by Galsworthy, MJ. Hosted by the Institute of Biomedical Informatics (IBMI), Faculty of Medicine, University of Ljubljana,
Slovenia. URL: www.medsum.info.
Table 1
Market overview on micro-CT: manufacturers and their current products.
Company Web site Products
Biomedical Imaging Research (BIR) http://www.bio-imaging.com ACTIS 150/90 Desktop
Varian Inc. ACTIS 150/130 Desktop
Lincolnshire, IL, USA ACTIS 200/225 Desktop
Bioscan, Inc.
4590 MacArthur Blvd., NW, USAWashington, DC 20007
http://www.bioscan.com NanoSPECT/CT
NanoPET/CT
Echo Medical Systems
Houston, TX, USA
http://www.echomri.com LaTheta LCT-200
LaTheta LCT-100A
Gamma Medica-ldeas, Inc.
Northridge, CA, USA
http://www.gm-ideas.com FLEX Triumph
GE Medical Systems
Waukesha, Wl, USA
http://www.gehealthcare.com eXplore Locus
eXplore Locus SP
eXplore CT 120
eXplore Vista PET/CT
Triumph
SCANCO Medical AG
Brüttisellen, Switzerland
http://www.scanco.ch viva CT 75
viva CT 40
extremCT
lCT 35
lCT 40
lCT 80
Siemens AG
Erlangen, Germany
http://www.medical.siemens.com Inveon Micro CT
SkyScan
Kontich, Belgium
http://www.skyscan.be SkyScan 1076 in-vivo
SkyScan 1178 high-throughput
SkyScan 1172 high-resolution
SkyScan 1174 compact
Stratec Medizintechnik GmbH
Pforzheim, Germany
http://www.stratec-med.com XCT Research SA/SA+
XCT 3000 Research M/M+
Orthometrix Inc.
Naples, FL, USA
http://www.orthometrix.net XCT FAN Beam l-Scope
VAMP GmbH
Erlangen, Germany
http://www.vamp-gmbh.de TomoScope 30s
TomoScope 30s+
TomoScope Duo
YXLON International GmbH
Garbsen, Germany
http://www.yxlon.com Yxlon Y.Fox lCT
S.J. Schambach et al./ Methods 50 (2010) 2–13 3
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 3/12
In the literature, both customized non-destructive testing lCTs
[16] and custom-made lCTs [3,17] adapted for the necessities of animal research are described. The following paragraphs will detail
some of the various lCT scanner designs and discuss the advanta-
ges and limitations of this form of imaging technology.
2.1. lCT setup
2.1.1. Construction principles
Generally, there are two different construction principles with
respect to lCT scanners. The first construction principle involves
scanners with an X-ray detector and radiation source mounted
on a gantry that is rotated around the examined object (Fig. 2A).
In these scanners the source–detector distance (SDD) is construc-
tion-conditioned in most marketed tomographs (with a few excep-
tion where SDD can be adjusted), with a magnification level for theobject in the course of the beam. In these systems the achievable
geometrical resolution depends mainly on the pixel pitch and
matrix size of the detector used, as well as the focusing mode of
the X-ray tube used. Accordingly, scanners are named ‘‘mini-CT”
because the setup is analogous to clinical CTs.
The second construction principle, which is found mostly in
ex vivo specimen lCT scanners and custom built systems, is out-
lined in Fig. 2B. In this design the examined object is rotated within
the light path. The setup allows the free adjustment of source–
object distance (SOD) and object–detector distance (ODD), thereby
allowing SDD adjustment. Free adjustment of SOD and ODD facili-
tates optimization of the geometric magnification level, depending
on the signal-to-noise ratio (SNR) and the penumbra blurring [15].
Thus, for small fields of view, higher maximal resolution can be
achieved as compared to a conventional setup. In these systems,
the object can be rotated horizontally [16] or vertically [17]
orthogonal to the ray path. However, one drawback is the necessity
to fix the examined animal during rotation around its own axis, to
prevent movement blurring in the resulting datasets.
2.1.2. X-ray tubes
The availability of a variety of lCT scanners using a range of
X-ray tube technologies presents both advantages and disadvan-
tages. The majority of marketed scanners use nano- or microfocus
X-ray tubes with transmission targets. In these X-ray tubes an
electron beam is produced on the tip of a hairpin tungsten filament
and focused by several magnetic lenses onto a focal spot of
1–10 lm on a transmission target (Fig. 3).
Transmission targets typically have a thin layer of tungsten
(about 100 lm) electroplated or vapor-deposited on a carrying
material with low atomic number and high thermal conductance
such as beryllium or chemical vapor depositioned (CVD)-diamonds(e.g. Diamond Materials GmbH, Freiburg, Germany). On the outside
of the transmission targets a cone-shaped beam of braking radia-
tion (depending on tube voltage), and characteristic radiation
(depending on the target material) is produced which irradiates
onto a digital X-ray detector.
In contrast to this, in conventional clinical CT or three-dimen-
sional (3D)-rotation angiography systems, X-ray tubes with reflec-
tion targets and focal spot sizes of a minimum of 300lm are
typically used. If tubes of this dimension are used, as per the pro-
tocol of Badea et al. [17], then a low magnification must be used in
order to minimize penumbra blurring caused by the large focal
spot. Reflection target X-ray tubes convert the irradiated electron
energy less efficiently to X-ray photons than transmission target
tubes [18]. On the other hand, reflection targets can absorb moreheat energy without damage because they have a thicker tungsten
anode compared to the tungsten layer on transmission targets.
Therefore, higher energy electron beams can be used in reflection
anode tubes, leading to higher photon flux in these tubes [17].
Reflection anode tubes typically generate power in the range of
kilowatts, whereas microfocus transmission tubes operate in the
range of watts. In 1994, Flynn et al. stated that the output power
of a microfocus tube would generally follow P max = 1.4( x)0.88
[W/lm] where x equals the focal spot size in lm [19]. Due to the
use of advanced target materials, such as CVD-diamonds that have
Fig. 2. Different lCT architectures. In thedesign illustrated under (A), theexamined
object is placed still in the center of the setup and a gantry carrying detector and X-
ray source is rotated around it. The geometrical magnification factor is fixed
structurally by the defined SDD. In the setup outlined in (B), the object is rotated in
the course of the beam and can be freely positioned between detector and source,
which allows for the adjustment of the magnification level.
Fig. 3. (A) Transmission target X-ray tube of the Y.Fox lCT. (B) Sketch of the inside configuration of transmission tubes with the electron beam exiting a hairpin filament
(triangle) that is focused viamagneticlenses(gray bars) on thetransmission target(1) . (C) Reflection anode X-ray tube with rotating anode in closed design and(D) sketchof
a reflecting anode X-ray tube design with electrons exiting from a curled heating cathode and the electrons accelerated onto a reflection target (1). (Image source of (C) and(D): Wikipedia Commons.)
4 S.J. Schambach et al./ Methods 50 (2010) 2–13
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 4/12
an extremely high warmth conductance, current microfocus tubes
far exceed these conditions. For example, the Y.Fox lCT can gener-
ate up to 13 W of power with a focal spot size of 1 lm and is lim-
ited only by the maximal power output of the voltage generator,
where a maximal output of 1.34 W would be calculated for the cor-
responding focal spot size.
2.1.3. Beam geometry
Differences between system types also exist in terms of beam
geometry. Using a fan beam X-ray source a 3D dataset is acquired
either plane by plane (incrementally) or under continuous feed
motion of source against object (spiral-CT) [20] via a line detector.
An alternative to fan beam CT is the so-called cone beam CT that
was first implemented by Feldkamp et al. [46,122] (also known
as volume-CT). These principles are outlined schematically in
Fig. 4. In cone beam lCT small objects are captured completely
in one rotation which considerably speeds up imaging compared
to incremental or spiral scanning. Cone beam lCT also facilitates
cardiac and general thoracic imaging in small rodents with faster
physiological heart and breathing movements. Relinquishing therotation of the object, these systems also have the capability to
perform 2D measurements such as conventional X-ray, or digital
subtraction angiography (DSA) due to the large field of view cover-
age by the cone-shaped beam (Figs. 4B and 6).
2.1.4. X-ray detectors
To record the desired image data in commercial lCT systems,
most often charge-coupled device (CCD) photodetectors coupled
to a scintillator by tapered glass fibers are used [21]. These detec-
tors do not have any readout electronics covering the photosensi-
tive layer, resulting in a so-called fill factor of 100% and leading to a
high signal yield. However, readout is facilitated by shift register
readout with one signal output over the entire detector, leading
to longer readout times.The closest competitors are active matrix flat panel imagers
(AMFPI) that consist of an array of photodiodes connected by a ma-
trix of thin film transistors (TFTs) such as TFT liquid crystal image
displays. The pixels are read out row by row speeding up readout
times compared to CCD detectors [22], but the TFTs cover up to
50% of the photosensitive layers (fill factor6 50%) and reduce light
sensitivity.
2.1.5. Contrast generation and enhancement possibilities
Contrast in lCT images is produced most of the time by atten-
uation of X-rays in the examined sample [23]. Below 25 keV, X-ray
attenuation mainly takes place by the photoelectric effect, and the
resulting X-ray energy after attenuation is anti-proportional to the
third power of the atomic number. At higher X-ray energy levelsCompton scattering is the main cause of attenuation leading to
an anti-proportional dependency of attenuated X-ray energy to
the atomic number divided by 2. Accordingly, at low X-ray energy
the contrast between different tissues containing a mixture of ele-
ments with different atomic numbers is much higher than the
threshold of 25 keV [24].
X-ray contrast can be increased by various physical methods
such as K-edge subtraction [25–28], X-ray phase delay contrast
[29–32] and X-ray scatter contrast [33–36]. With these methods,
up to 15 lg iodine/cm3 [37], 250 lg iodine/cm3, and 6 mg iodine/
cm3 [38] can be detected, respectively, compared to 10 mg
iodine/cm3 without using the above-mentioned physical methods
for contrast enhancement.
lCT offers good image quality for objects consisting of elements
with high atomic number such as bones, but possesses relatively
weak soft tissue contrast. Therefore when imaging soft tissues,
the administration of a contrast agent frequently is desirable and
sometimes indispensable.
With regard to circulation time, two different types of contrast
agents are generally available for in vivo imaging: (a) a conven-
tional iodinated contrast medium that is eliminated from the blood
immediately by the kidneys or (b) a blood-pool contrast agent,
with a higher blood-pool half-life.
In 2004 Ritman [13] noted that up until then it was not possible
to image a bolus of an injected contrast agent within a live rodent
even with fast synchrotron scanning lCTs. Since then, faster X-ray
detectors have been developed to read out between 30 and 60 fps
and allow for angiography using conventional contrast agents.
Water-soluble, non-ionic iodine-based contrast agents are gener-
ally used in humans and are eliminated immediately from the
blood after intravenous injection. However, in humans, it is also
possible to acquire images in the first pass of the contrast agent,
due to a relatively long circulation time, as compared to the rodent,
and a sufficient temporal resolution found in clinical CTs. The rapid
elimination of non-ionic iodine-based contrast agents from the
blood restricts the administration of this contrast agent to a small
percentage of the total human blood volume. Due to very short cir-
culation times in rodents [39] a first-pass effect cannot be used asin humans. The contrast agent is eliminated from the blood within
seconds, leading to a need for continuous injection of a contrast
agent above the renal elimination rate, and throughout the scan-
ning time [40]. Previous studies have demonstrated sufficient ves-
sel contrast with an injection volume of 350 ll/30 g mouse, which
was hemodynamically well tolerated by all animals [16]. However,
a reduction of injection volume is desirable to minimize the
physiological alterations in the model organisms by intravenous
volume burden.
Experimental setups with longer scan times can benefit from
blood-pool contrast agents, consisting of molecules that bind to
plasma proteins [41] or macromolecules that are eliminated at a
slower rate because of the sheer size of these molecules [42].
Blood-pool contrast agents allow scanning times that last up toseveral hours after a single administration, whereas the level of
contrast is generally lower than that achieved by conventional con-
trast agents. Lower levels of contrast limit the use of blood-pool
contrast agents to larger vascular structures, but allow repeated
measurement of the same animal and evaluation of eventual vessel
changes without the need for repeated contrast agent administra-
tion and volume burden. The potential for longer scanning times
helps maximize the SNR, which is dependent on the inverse loga-
rithm of the reconstruction input image number. Longer scanning
times produce sharper datasets at the cost of a high radiation dose
for the examined animal.
Newer liposome-based blood-pool contrast agents with a high-
er iodine concentration of greater than 100 mg/ml and a longer
blood-pool half-life offer improved detail perceptibility particu-larly in small anatomical structures [43].
Fig. 4. Principles of fan beam (A) and cone beam (B) lCT: (1) X-ray source, (2)
beam, (3) rotatedobject in the course of beam in longitudinal motion (A) or in fixed
z -position (B), and (4) line (A) or area (B) detector.
S.J. Schambach et al./ Methods 50 (2010) 2–13 5
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 5/12
Another novel liposome-based contrast agent tested by Montet
et al., with an iodine content of 70 mg/ml, allows for the measure-
ment of hepatic metastasis of around 250 lm in size, as well as
delineation of the liver and spleen. Furthermore, the iodinated
liposomes have been found to be a suitable contrast agent for
vascular structures [44]. More recently, multimodal contrast
agents are being tested and gaining considerable attention as they
contain both iodine and gadolinium and can therefore be used
either in CT or in MRI [15,45].
3. Application of lCT in experimental imaging
3.1. Osseous structures
The first advances in lCT technique were mainly driven by
imaging needs for the evaluation of bone anatomy and density
[46,47]. These publications have investigated bone density [48];
osteogenesis [49]; ovariectomy [50] and osteoporosis [51]; bone
resorption [52]; bone remodeling [53]; bone regeneration [54]
and fracture healing [55]; bone neoplasm [56] and biocompatible
materials [57]; and many more topics. Various reviews have also
addressed the use of lCT in the evaluation of pathological changesin bone structure [48,58–60].
The high X-ray density of osseous structures allows the precise
lCT-based evaluation of stereology, volume, and trabecular archi-
tecture of bones at micrometer resolution [61]. For the volumetric
estimation of bone density, ex vivo lCT is described as the method
of choice [62]. Furthermore, the non-invasive quality of lCT allows
for the observation of bone structure before and after exposure to
mechanical stress under experimental conditions [12]. An isotropic
resolution of about 50 lm is described as sufficient to evaluate
changes correlated with osteoarthritis [63] and other bone-remod-
eling processes in rats in vivo [64]. In studies of osseous disease,
differences in trabecular structure and mineralization density,
which can have an impact on experimental setup and data collec-
tion, were compared between different mouse strains in vivo [61].
The results showed a higher degree of mineralization in C3H/HeH
mice compared to C57BL/6 mice at identical body weight and body
size in concordance with earlier ex vivo studies [65]. According to
this study, lCT is capable of detecting the impact of illnesses and
therapeutic interventions on bone density and structure [61]. To
provide an example for imaging of osseous structures, Fig. 5 shows
a murine proximal femur volume rendering of a dataset acquired
via lCT ex vivo, where we compared a fast (A: scan time 40 s)
and a slow scan mode (B: scan time 20 min).
Next to the evaluation of changes in trabecular structures, the
quantification of osteopathologic processes is also very important.
For this purpose, lCT is used in fundamental research in oncology
for the quantification of bone metastases [66]. Besides osteolytic
alterations or processes, an increase in bone density, such as
osteopetrosis, manifested in op/op knock out mice, can also be
assessed using lCT [67].
3.2. Vascular structures
To date, approximately 60 articles have been published on the
evaluation of vasculature in small animals via lCT. Starting in
the late 1990s, topics such as angiogenesis [40,68,69] and neovas-
cularization [70] were studied. In addition, the vasculature of mis-
cellaneous organ systems, both in healthy and in diseased
conditions, have been evaluated in terms of renal vasculature
[71–74]; hepatic vasculature [75] and portal hypertension [76];
Fig. 5. Volume rendering of a murine femur dataset, acquired ex vivo with a continuous scan mode lasting 40 s (A) and with an incremental scan mode having 20 minacquisition time (B) using a volume-CT.
Fig. 6. Digital subtraction angiography of the cerebral arteries after superselective
catheterization of the common carotid artery in a rat (cranio-caudal view).
6 S.J. Schambach et al./ Methods 50 (2010) 2–13
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 6/12
cerebral vasculature [16,77]; coronary arteries [78]; and ocular
vasculature [79].
The in vivo evaluation of vascular structures in small animals
via lCT is only possible by means of contrast agent administra-
tion, which is similar to CT angiography (CTA) in humans. The
earlier, relatively slow lCT scanners with scanning times lasting
up to hours allowed for vessel analysis in rodents after sacrifice
of the animal [13] using perfusion with radiopaque polymerizing
substances [77,80] or shock freezing the contrast agent perfused
specimen [81]. In ex vivo studies there is no need for X-ray dose
reduction, and no anesthesia-dependent limitations in scanning
time, therefore SNR can be maximized by long scans with high
photon flux, resulting in submicron resolution [80]. Recent ad-
vances in X-ray detector technology, leading to faster readout
times, lager pixel matrices, higher X-ray sensitivity, and higher
SNR, allow examination times of less than a minute [16,40].
These advances have in turn lead to a reduction of movement
artifacts, anesthesia incidents, applied radiation dose, and the
use of conventional contrast agents with lCT angiography. Fur-
thermore, not only does using systems with cone-beam geome-
try allow computed tomography to be performed, but also
digital subtraction angiography is possible as demonstrated in
Fig. 6.
Actual examples for high-resolution 3D vessel imaging in vivo
using conventional contrast agents were reported by Kiessling
et al. and Schambach et al., when they imaged tumor supplying
vessels at resolutions of 50 lm [40] and cerebrovascular structures
in mice at 16 lm resolution, respectively [16]. Figs. 7 and 8 show
in vivo datasets of intracranial and extracranial vessels of a mouse
using a conventional contrast agent (Iomerone 300, Bracco Altana)
with 40 s scanning time. These datasets allow not only the analysis
of anatomical differences in brain vasculature between mouse
strains, but also the evaluation of acute vessel diameter alterations
between hypoxia and normoxia.
The use of so-called blood-pool contrast agents for imaging of
vascular structures as well as parenchymal organs in small animals
is well described in the literature [82–91]. Various thoracic and
abdominal murine vascular structures acquired via lCT in vivo
are displayed in Fig. 9 using the liposomal contrast agent Fenestra
VC.
However, some vascular pathologies can be detected without
the application of a contrast agent. For example, Persy et al. evalu-
ated the degree of aortic calcification in a model of chronic renal
failure of rats in vivo without contrast agents anddrew conclusions
about the development and the degree of renal failure in these
animals [92].
Fig. 7. Maximum intensity projections (MIP, A–C) and volume rendering (D) of murine cerebral vessel datasets, acquired with a lCT in vivo using a conventional contrast
agent and bolus technique. Image (A) shows the passage of the internal cerebral artery (ICA) through the skull base and the circle of Willis incorporating the middle cerebral
artery (MCA) and the anterior cerebral artery (ACA), presented in a curved-MIP. In image (B) a transversal view of the circle of Willis of a BALB/c mouse with prominent
posterior communicating artery (PcomA) between posterior cerebral artery (PCA) and superior cerebellar artery (SCA) is shown. Image (C) represents a sagittal view of
murine brain vessels with small branches of the azygos of the pericalosal artery (azPA) visible. The azPA is supplied in mice by a unification of both ACA called azygos of the
ACA(azACA). (D) shows a volumerendering of external cranial vessels of a mouse with arteries such as the commoncarotid a. (CCA), external carotid a. (ECA),internal carotida. (ICA), stapedial a. (SA), occipital a. (OA), caudal auricular a. (CAA), superficial temporal a. (STA), facial a. (FA), and lingual a. (LA) visible.
S.J. Schambach et al./ Methods 50 (2010) 2–13 7
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 7/12
3.3. Cardiothoracic imaging
3.3.1. Gating strategies
In this section, for practical reasons gating strategies are dis-
cussed in terms of how they are applied in cardiothoracic imaging,
rather than in a separate paragraph in Section 2.
High resolution cardiothoracic imaging of rats and mice poses a
challenge to the researcher because of the relatively small size of
the animals and because of the high respiratory and cardiac
frequencies. The heart rate of mice ranges between 400 and 600
beats per minute (bpm), whereas the heart rate of rats is a little
bit lower at 250–400 bpm. At rest, the breathing frequency of mice
is about 200 comparedto about 70–110 respiratory cycles per min-
ute for rats. Depending on the depth and kind of anesthesia, the
heart and breathing rate can be lowered. The ultimate strategy to
control breathing during experimentation requires intubation
and controlled ventilation of laboratory animals as per the protocol
described by Namati et al. [93].
Cardiopulmonary imaging in rodents using lCT is challenging
due to physiological motion of the cardiac and respiratory systems.
To achieve high resolution imaging of lung structures, while avoid-
ing intubation, or for cardiac imaging in vivo, the application of
gating strategies is necessary. The blurring that is caused by the
physiological movement of the diaphragm or the heart can be
effectively reduced with cardio-respiratory gating, which can be
done either prospectively or retrospectively. In this correlation
prospective strategies are to be distinguished from retrospective
approaches, in that the latter can be subdivided into extrinsic
and intrinsic gating methods.
In prospective gating, image acquisition takes place only during
defined breathing or heart phases and requires direct synchroniza-
tion between lCT and physiologic parameters (ECG and/or breath-
ing signal) of the animal. Prospective gating is already
implemented in the majority of lCT systems by Bioscan, GMI,
GE, Siemens, SkyScan and VAMP. An example for the successful
adoption of prospective gating is described in a publication by
Bartling et al., in which a significant improvement of image quality
was attained via prospective gating in lCT scans of rats, rabbits
and mice [94].
Retrospective gating can be subdivided into extrinsic and intrin-
sic gating methods. Extrinsic retrospective gating requires syn-
chronization of physiological data acquired during the scanning
(e.g. ECG or respiratory movements) and the imaging data. Theassignment of the individual image projections to respiration or
heart phase is done after data acquisition [87]. However, to assign
time-points to recorded image data and physiological data for syn-
chronization requires direct hardware access to both the X-ray
detectors used and frame grabber.
Intrinsic retrospective gating can be implemented without
hardware modifications on any lCT scanner, as long as the physi-
ological signals of heartbeat and diaphragm movement can be
sampled from the image projections itself. Of note, however, com-
plex post-processing algorithms are generally needed. Various
examination protocols are possible in this context. On the one
hand, it is possible to acquire a set of images from a single angle
position to identify the different breathing/heart phases in these
images and repeat this process from several angles. The secondpossibility consists of continuous image acquisition with a high
frame rate at a constant rotation speed and the rearrangement of
single projections of the resulting dataset according to the heart/
breathing phase. In this setup, generally several scans are
performed in a row [94], then the images acquired during breath-
ing excursions are discarded and the remaining images are
grouped by heart phase to new raw datasets. Fig. 10 shows a data-
set acquired with a volume-CT scanner using this last method to
determine left ventricular function, myocardial thickness, and lung
volume in mice, to name a few.
3.3.2. Thoracic imaging
The successful implementation of thoracic lCT allows the
detection of pulmonary lesions in a murine adenocarcinoma model[2]. In this study, Li et al. were able to detect lung tumors P1 mm
Fig. 8. Images (A + B) show volume rendering datasets of the circle of Willis of a
BALB/c mouse, (C + D) of a C57BL/6 mouse with visible anterior cerebral a. (ACA),
middle cerebral a. (MCA), internal carotid a. (ICA), posterior cerebral a. (PCA),
superior cerebellar a. (SCA), posterior communicating a. (PcomA) and basilar artery
(BA). In BALB/c mice the posterior circulation is mainly supplied over the PcomA,
whereas in C57BL/6 mice the related territory is supplied by the BA and SCA, which
can lead to different infarction territories in murine stroke models corresponding tothe difference shown in vessel anatomy.
Fig. 9. Volume rendering of ungated murine datasets using the blood-pool contrast
agent Fenestra VC
. (A) shows abdominal vasculature and (B) shows thoracicvasculature. Veins and arteries are equally visible in both datasets.
8 S.J. Schambach et al./ Methods 50 (2010) 2–13
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 8/12
in diameter in unenhanced scans. Even though a sharp cutoff for
the tumor border could not be defined, it was still possible to dis-
tinguish the lesion from abutting vascular structures with the help
of other slice orientations. In a more recent study, it was possible todelineate pulmonary tumors of 0.85 mm in vivo in a nude mouse
model [95]. One of the latest in vivo studies of mice attained de-
tailed perceptibility of lung tumors of 500 lm using respiratory
gating [96].
Other studies describe the capability of lCT for the monitoring
of lung fibrosis in mice after intratracheal administration of
bleomycin [97]. The fibrotic lung segments present themselves
with a rising compaction of lung parenchyma. In another model,
emphysema was produced in mice via tracheal instillation of pan-
creas elastase and the typical changes in lung architecture could be
detected [98]. The differences in regional lung ventilation were
non-invasively imaged in Wistar rats via xenon gas inhalation
lCT [99].
3.3.3. Cardiac imaging
Hypercholesterolemia and arteriosclerosis can produce oxida-
tive stress, a dysfunction of coronary arteries and myocardial
ischemia, which accompany the expression of growth factors and
can lead to myocardial neovascularization. Using high-resolution
lCT ex vivo, Zhu et al. demonstrated that it was possible to image
the related changes in myocardial micro-vasculature in swine
[100]. In this quantitative analysis, the subendocardial spatial den-
sity of microvessels with a diameter under 200 lm, at an imaging
resolution of 40 lm was significantly higher in swine suffering
from hypercholesterolemia compared to controls.
For in vivo imaging of the heart in rat and mice, projections
must be assigned not only to different breathing phases, but also
to higher frequency heart actions. Cardiac gating in small animalshas been successfully described in the literature in terms of the
application of prospective and retrospective gating techniques
[84,101]. Using pulsed X-ray tubes as done by Badea et al. [84], a
10 ms time window per frame is possible. Temporal resolution of
a dedicated X-ray detector should be high enough to display oneheart cycle in a set of images. Most detectors in use do not exceed
a frame rate of 30 fps. The implication here is that at frame rates of
30 fps the end-diastolic and end-systolic phase can be identified in
animals with a heart rate of up to 300 bpm. Images that we pro-
duced under such conditions via retrospective intrinsic gating in
mice are presented in Fig. 10.
To summarize, in vivo cardio-CT is well established as a valu-
able method of non-invasive investigation in heart pathology mod-
els as well as for the phenotyping of genetically modified small
animals.
3.4. Imaging of abdominal organs
lCT imaging of abdominal organs in small animals is increas-ingly used as the availability of scanners has greatly improved. In
this context, the examination of parenchymatous organs such as
liver, spleen and kidney is of high interest, whereas most studies
focus on oncological problems and the investigation of the kidney
function.
3.4.1. lCT of parenchymatous upper abdominal organs
Relatively long scanning times, in the range of minutes, motion
blurring caused by breathing movements, and the small blood
volume of animals make the evaluation of subphrenic abdominal
organs, like the liver, difficult when using conventional iodinated
contrast agents. As a result, for the imaging of parenchymatous
upper abdominal organs, liver-specific contrast agents are used
on a regular basis. For example, the iodinated contrast agentDHOG (1,3-bis-[7-(3-amino-2,4,6-triiodophenyl)-heptanoyl]-2-oleoyl
Fig. 10. Datasets of the murine heart after retrospective intrinsic gating. Images (A + B) show a short axis view in MIP mode with (A) representing end-systolic and
(B) end-diastolic phase. In (D) a long axis view of the heart in end-diastolic phase with clearly distinguishable left ventricle (LV), right ventricle, atriums and aortic arch (AA).
(C) shows a volume rendering with pulmonal tissue visible in light blue and trachea as well as bronchioli in red. (Intrinsic gating in cooperation with Q. Xie, Experimental
Radiation Oncology, Medical Faculty Mannheim, and S. Bartling, German Cancer Research Centre (DKFZ).)
S.J. Schambach et al./ Methods 50 (2010) 2–13 9
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 9/12
glycerol; Fenestra LC; Art Technologies, San Diego, CA), which is
internalized by hepatocytes via the ApoE-receptor and leads to a
relevant enhancement of liver contrast after 1–2 h [86,88,102].
DHOG allows for the delineation between healthy, enhancing
hepatocytes and non-enhancing neoplasmatic cells [103] and thusthe non-invasive quantification of liver metastases. The burden-
some ‘‘second look” procedure used in older experimental setups
can, therefore, be abandoned. Furthermore, liposomes are ingested
by phagocytic cells in the spleen, and the red pulp which is
populated mostly by erythrocytes and macrophages. This ingestion
creates a hyper-dense imaging effect [44]. As the half-life of DHOG
can range from several days up to 2 weeks, it may offer the possi-
bility to image the tumor growth repeatedly without the need for
repeated injections, or at least to reduce the dose of the adminis-
tered contrast agent in repeated scans.
The smallest neoplasms that could be confirmed in a respira-
tory-gated DHOG enhanced lCT scan have been reported to be in
the range of 250–300 lm [44]. However, Fig. 11 shows a murine
liver interspersed with metastases with diameters of less than100 lm scanned about 2.5 h after i.v. administration of 400ll
Fenstra LC.
3.4.2. lCT of the kidney
Up to 10 years ago it was difficult to geometrically relate the 3D
anatomical complexity of renal vasculature with the related tubu-
lar sections of the nephron because of methodological limitations.
A recent review on the use of lCT for the evaluation of renal micro-
structural changes addresses the paucity of imaging publications
citing only 16 articles [104]. Most of the cited articles evaluated
anatomy ex vivo to compare 3D anatomy to histological slices from
human specimens.
Likewise, only a few publications describe the use of lCT to
evaluate renal structures in animals. Using lCT, Fortepiani et al.showed that the blockage of nitrogen-oxide synthesis (an experi-
mental approach to elevate blood pressure) leads to a reduction
in renal blood flow and that administration of an AT1-receptor
antagonist inhibited the effects of this blockage. To evaluate vessel
changes, the kidneys were perfusion fixed in situ and perfused
with radiopaque silicon. Furthermore, an elevated heterogeneity
of glomerular volumes in the renal cortex of rats suffering from
diabetic nephropathy was confirmed via lCT [105].
Even fewer studies on the use of lCT for in vivo imaging in
small animals exist. A study by Almajdub et al. describes an
in vivo imaging procedure for mouse kidney anatomy evaluation
using contrast-enhanced high-resolution lCT [106]. They demon-
strate that contrast-enhanced lCT enables accurate in vivo mea-
surement of kidney volume, length and thickness in mice andreport reference parameters for four strains.
Other authors applied lCT of the kidneys in small animals to
evaluate liposomal blood-pool contrast agents by ruling out renal
excretion [43,89].
As an example of renal in vivo lCT imaging, Fig. 12 presents ascan of the renal pelvis and the ureter of a mouse after contrast
agent administration is presented.
3.4.3. Gastrointestinal tract
Colon carcinoma is one of the most frequent tumors in Western
society. Various animal models exist to mimic this disease andlCT
is applied successfully in this research field as well. For example, in
a study by Durkee et al., virtual colonoscopy was performed anal-
ogous to the examination of colon polyps in humans. After nega-
tive contrasting by air insufflation in APC-min-mice, colon polyps
could be visualized and measured [107]. In a subsequent histolog-
ical examination, a high correlation between identified tumor vol-
umes via lCT and histology was found. Using various positivecontrast agents, it was also possible to image spontaneously
Fig. 11. MIP of the murine liver contrasted with Fenestra LC in transversal (A) and coronal (B) view (1200 projections, scan time 40 s, Field of View 3.5 3.5 cm). White
arrows point to hypo dense round structures with a diameter of 100 lm, in the sense of metastases.
Fig. 12. Volume rendering of lCT scan of a C57BL/6 mouse in vivo after injection of
the blood-pool contrast agent Fenestra VC, acquired without cardiac or pulmonary
gating. The numbers refer to the abdominal vena cava (1), the renal vein (2), right
iliac vein (3), spleen (4), kidney (5), and the ureter (6).
10 S.J. Schambach et al./ Methods 50 (2010) 2–13
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 10/12
emerging colon polyps in Cdx2+/ mice, as well as in mice bearing
azoxymethane-induced polyps [108]. Fig. 13 shows a 3D virtual
colonoscopy dataset of a mouse, as well as a two-dimensional
X-ray with a double-contrasted colon performed on a volume-CT.
3.4.4. Quantification of body and organ fat content
The estimation of body and organ fat content is important be-cause of a high quantity of metabolic syndromes with a related
alteration in the percentage of body fat. Although various other
methods allow for the quantification of body fat, lCT offers a
non-invasive and elegant possibility to quantify not only body fat
volume, but the degree of fatty degeneration of distinct organs as
well [109].
3.5. lCT of cerebral structures
Several ex vivo studies analyzing cerebral microcirculation and
anatomical details of small rodents have been performed, as they
allow longer scans on samples to achieve a higher SNR and higher
spatial resolution after appropriate treatment of the tissue [110–
113]. For example, it was possible to distinguish the white andgraymatter, as well as tumor tissue in mouse brain samples after they
were allowed to absorb iodinated contrast agent [114].
While cranial CT is a routine method in clinical CNS examina-
tions, only a few publications describe the use of lCT for CNS
examination in small animals. In a study by Newcomb et al., the
size of intracerebrally growing invasive gliomas in mice was mea-
sured using a clinical CT scanner [115]. The image quality was suf-
ficient to see the contrast-enhancing tumor parts with moderate
resolution. Engelhorn et al. were the first to report the use of a
lCT for double dose contrast-enhanced in vivo imaging of xeno-
grafted rat brain glioma [116]. They found a good correlation of
lCT- and 3.0T clinical MRI-derived tumor volumes compared to
histology. Another in vivo study was conducted by Balvay et al.
who examined the microcirculation of gliomas in Wistar rats viadynamic contrast-enhancedlCT, although they used a synchrotron
radiation source [117]. To our knowledge, to date there are no pub-
lications exploring intracranial tumors in mice using lCT in an
in vivo setting.
4. Concluding remarks
4.1. Summary and future directions
With the increased availability and user-friendliness of lCT,
there has been an increased opportunity for preclinical research.
Simple operability, fast scanning protocols, a significantly higher
temporal and spatial resolution, plus lower acquisition costs and
maintenance, all prove beneficial compared to high-field small ani-mal MRI. However, with a view to MRI, lCT can also be seen as a
complementary, additional and adjuvant technique rather than a
competitive one. Disadvantages in comparison with MRI, are the
lower soft tissue contrast, and the particularly high radiation dose.
For example, measurements with thermoluminescent dosimeters
implanted in mice yield a radiation dose of 10–50 cGy per CT scan
[10], whereas with very small source–object distance dose levels of
up to 5 Gy with dermal depilation after 2–3 weeks are reported[16]. Therefore, the cumulative dose can be very high, especially
in longitudinal studies, as such the radiation associated biological
effects depend not only on radiation dose but also on the applica-
tion time frame. As a result, the radiation dose needs to be consid-
ered in the planning of oncological in vivo studies using lCT. A
cumulative whole body radiation dose of 50 cGy within 7 days
leads to a significant increase in the incidence of ovarian cancer
in BALB/c mice [118], and a dose of 1 cGy to a significant reduction
of tumor volume in lymphoma mouse models [119]. A whole body
dose of 50 cGy decelerates the progression of diabetes type I in
non-obese diabetic mice [120]. Based on these findings, the chal-
lenge in lCT imaging will be the acquisition of valuable datasets
at much lower radiation doses. These drawbacks can be minimized
using novel contrast agents, and up-to-date low-dose l
CTs [121]
with newer, more sensitive detectors and intelligent scanning
protocols.
Along with an increase in the employment of lCT, the growing
availability of PET-CTs can be expected from various companies
such as Bioscan, GE, GMI or Siemens (for details, see the homepag-
es listed in Table 1). PET offers extremely high sensitivity in the
detection of targeted neoplasms at a low resolution, which is coun-
terbalanced by an image overlay of a high-resolutionlCT scan that
is acquired in the same session. The fusion of lCT and fluorescence
imaging could be of interest as well and would obviate the use of
radioactive tracers. However, this fusion could only be applied to
small animals such as mice or for superficial processes because
of the low light intensity and tissue penetration of fluorescent
light. The combination of two procedures like PET or fluorescence
imaging and CT does not necessarily mean the acquisition of a
new combined scanner but can be achieved via a subsequent fu-
sion of datasets recorded with separate scanners.
Starting from an ongoing optimization of scanner equipment
such as more sensitive detectors with even higher temporal and
spatial resolution, reduced radiation doses, and further improve-
ments in software protocols, a long-term increase in the use of
lCT in preclinical research can be expected. Additionally, improve-
ments in gating algorithms and iterative low noise reconstruction
methods may lead to an increasing acceptance and availability of
lCT methods.
References
[1] E.L. Ritman, Nucl. Instrum. Methods Phys. Res. A 580 (2007) 968–970.
Fig. 13. (A) Raw data projection of a murine (APC (min)) virtual colonoscopy dataset acquired in vivo with double contrast, white arrows pointing at the colon polyps.
(B) Volume rendering of a similar dataset, showing the internal surface of a healthy colon.
S.J. Schambach et al./ Methods 50 (2010) 2–13 11
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 11/12
[2] D. Cavanaugh, E. Johnson, R.E. Price, J. Kurie, E.L. Travis, D.D. Cody, Mol.
Imaging 3 (2004) 55–62.
[3] M.A. Kujoory, B.J. Hillman, H.H. Barrett, Invest. Radiol. 15 (1980) 148–154.
[4] T. Sato, O. Chieda, Y. Yamakashi, M. Tsubouchi, Appl. Opt. 20 (1981) 3880–
3883.
[5] J.C. Elliott, S.D. Dover, J. Microsc. 126 (2) (1982).
[6] P. Burstein, P.J. Bjorkholm, R.C. Chase, F.H. Seguin, Nucl. Instrum. Methods
Phys. Res. 221 (1984) 207–212.
[7] C.V. Burton, Orthop. Clin. North Am. 14 (1983) 539–551.
[8] J.H. Kinney, Q.C. Johnson, M.C. Nichols, U. Bonse, R.A. Saroyan, et al., MRS Bull.
13 (1988).[9] U. Bonse, F. Busch, Prog. Biophys. Mol. Biol. 65 (1996) 133–169.
[10] M.J. Paulus, S.S. Gleason, S.J. Kennel, P.R. Hunsicker, D.K. Johnson, Neoplasia 2
(2000) 62–70.
[11] M.J. Paulus, S.S. Gleason, M.E. Easterly, C.J. Foltz, Lab. Anim. 30 (2001) 36–45.
[12] D.W. Holdsworth, M. Thornton, Trends Biotechnol. (2002).
[13] E.L. Ritman, Annu. Rev. Biomed. Eng. 6 (2004) 185–208.
[14] S.H. Bartling, W. Stiller, W. Semmler, F. Kiessling, Curr. Med. Imaging Rev. 3
(2007) 45–59.
[15] C.T. Badea, M. Drangova, D.W. Holdsworth, G.A. Johnson, Phys. Med. Biol. 53
(2008) R319–R350.
[16] S.J. Schambach, S. Bag, V. Steil, C. Isaza, L. Schilling, C. Groden, M.A.
Brockmann, Stroke 40 (2009) 1444–1450.
[17] C. Badea, L.W. Hedlund, G.A. Johnson, Med. Phys. 31 (2004) 3324–3329.
[18] D.J. Haas, Transmission x-ray tube, U.S. Patent 4,034,251, July 05, 1977.
[19] M.J. Flynn, S. Hames, D.A. Reimann, S.J. Wilderman, Nucl. Instrum. Methods
Phys. Res. A 352 (1994) 312–315.
[20] W.A. Kalender, P. Vock, A. Polacin, M. Soucek, Rontgenpraxis 43 (1990) 323–
330.
[21] C.M. Castelli, N.M. Allinson, K.J. Moon, D.L. Watson, Nucl. Instrum. Methods
Phys. Res. A 348 (1994) 649–653.
[22] G. De Geronimo, P. O’Connor, V. Radeka, B. Yu, Nucl. Instrum. Methods Phys.
Res. A 471 (2001) 192–199.
[23] M.E. Phelps, M.H. Gado, E.J. Hoffman, Radiology 117 (1975) 585–588.
[24] L. Grodzins, Nucl. Instrum. Methods Phys. Res. 206 (1983) 541–545.
[25] M. Torikoshi, T. Tsunoo, M. Sasaki, M. Endo, Y. Noda, Y. Noda, Y. Ohno, T.
Kohno, K. Hyodo, K. Uesugi, N. Yagi, Phys. Med. Biol. 48 (2003) 673–685.
[26] A. Sarnelli, A. Taibi, A. Tuffanelli, G. Baldazzi, D. Bollini, A.E. Cabal Rodriguez,
M. Gombia, F. Prino, L. Ramello, E. Tomassi, M. Gambaccini, Phys. Med. Biol.
49 (2004) 3291–3305.
[27] A. Sarnelli, H. Elleaume, A. Taibi, M. Gambaccini, A. Bravin, Phys. Med. Biol. 51
(2006) 4311–4328.
[28] B. Bewer, H. Zhang, Y. Zhu, L. Zhang, G.N. George, I.J. Pickering, D. Chapman,
Rev. Sci. Instrum. 79 (2008) 085102.
[29] A. Momose, Nucl. Instrum. Methods Phys. Res. 352 (1995) 622–628.
[30] F. Beckmann, U. Bonse, F. Busch, O. Gunnewig, J. Comput. Assist. Tomogr. 21
(1997) 539–553.
[31] J. Wu, T. Takeda, T.T. Lwin, A. Momose, N. Sunaguchi, T. Fukami, T. Yuasa, T.Akatsuka, Kidney Int. 75 (2009) 945–951.
[32] M. Bech, O. Bunk, C. David, R. Ruth, J. Rifkin, R. Loewen, R. Feidenhans’l, F.
Pfeiffer, J. Synchrotron Radiat. 16 (2009) 43–47.
[33] G. Harding, J. Kosanetzky, SPIE 671 (1986) 81–88.
[34] J.P. Schlomka, A. Harding, U. Stevendaal, M. Grass, G. Harding, SPIE 5030
(2003) 256–265.
[35] D.L. Batchelar, M.T. Davidson, W. Dabrowski, I.A. Cunningham, Med. Phys. 33
(2006) 904–915.
[36] S.R. Beath, I.A. Cunningham, Med. Phys. 36 (2009) 1839–1847.
[37] F.A. Dilmanian, Am. J. Physiol. Imaging 3/4 (1992).
[38] P. Duvauchelle, G. Peix, D. Babot, SPIE 3772 (1999) 268–278.
[39] J. Vogel, Am. J. Physiol. 273 (1997) H2520–H2527.
[40] F. Kiessling, S. Greschus, M.P. Lichy, M. Bock, C. Fink, S. Vosseler, J. Moll, M.M.
Mueller, N.E. Fusenig, H. Traupe, W. Semmler, Nat. Med. 10 (2004) 1133–
1138.
[41] X. Montet, M. Rajopadhye, R. Weissleder, ChemMedChem 1 (2006) 66–69.
[42] T. Barrett, H. Kobayashi, M. Brechbiel, P.L. Choyke, Eur. J. Radiol. 60 (2006)
353–366.
[43] S. Mukundan, K. Ghaghada, C. Badea, C. Kao, L. Hedlund, J. Provenzale, G. Johnson, E. Chen, R. Bellamkonda, A. Annapragada, Am. J. Roentgenol. 186
(2006) 300–307.
[44] X. Montet, C.M. Pastor, J.P. Vallee, C.D. Becker, A. Geissbuhler, D.R. Morel, P.
Meda, Invest. Radiol. 42 (2007) 652–658.
[45] J. Zheng, G. Perkins, A. Kirilova, C. Allen, D.A. Jaffray, Invest. Radiol. 41 (2006)
339–348, doi:10.1097/01.rli.0000186568.50265.64.
[46] L.A. Feldkamp, S.A. Goldstein, A.M. Parfitt, G. Jesion, M. Kleerekoper, J. Bone
Miner. Res. 4 (1989).
[47] P. Rüegsegger, B. Köller, Bone Miner. 25 (1994).
[48] K. Engelke, M. Karolczak, A. Lutz, U. Seibert, S. Schaller, W. Kalender,
Radiologe 39 (1999) 203–212.
[49] X.D. Liu, L.F. Deng, J. Wang, J. Qi, Q. Zhou, J.S. Wang, L. Wei, Y.P. Zhu, Zhonghua
Yi Xue Za Zhi 87 (2007) 3357–3361.
[50] Y. Gao, E. Luo, J. Hu, J. Xue, S. Zhu, J. Li, Bone 44 (2009) 225–232.
[51] B. Borah, G.J. Gross, T.E. Dufresne, T.S. Smith, M.D. Cockman, P.A. Chmielewski,
M.W. Lundy, J.R. Hartke, E.W. Sod, Anat. Rec. 265 (2001) 101–110.
[52] T.A. Freeman, P. Patel, J. Parvizi, V. Antoci Jr., I.M. Shapiro, J. Orthop. Res. 27
(2009) 673–679.
[53] J.U. Umoh, A.V. Sampaio, I. Welch, V. Pitelka, H.A. Goldberg, T.M. Underhill,
D.W. Holdsworth, Phys. Med. Biol. 54 (2009) 2147–2161.
[54] D.H. Kempen, L. Lu, A. Heijink, T.E. Hefferan, L.B. Creemers, A. Maran, M.J.
Yaszemski, W.J. Dhert, Biomaterials 30 (2009) 2816–2825.
[55] J.S. Nyman, S. Munoz, S. Jadhav, A. Mansour, T. Yoshii, G.R. Mundy, G.E.
Gutierrez, J. Biomech. 42 (2009) 891–897.
[56] L. Martiniova, M.S. Kotys, D. Thomasson, D. Schimel, E.W. Lai, M. Bernardo,
M.J. Merino, J.F. Powers, J. Ruzicka, R. Kvetnansky, P.L. Choyke, K. Pacak, J.
Magn. Reson. Imaging 29 (2009) 685–691.
[57] N. Tsukimura, M. Yamada, H. Aita, N. Hori, F. Yoshino, M. Chang-Il Lee, K.
Kimoto, A. Jewett, T. Ogawa, Biomaterials 30 (2009) 3378–3389.[58] R. Muller, P. Ruegsegger, Stud. Health Technol. Inform. 40 (1997) 61–79.
[59] T.M. Link, J.S. Bauer, Semin. Musculoskelet. Radiol. 6 (2002) 253–261.
[60] D. Chappard, M.F. Basle, E. Legrand, M. Audran, Morphologie 92 (2008) 162–
170.
[61] Y. Jiang,J. Zhao, D.L. White, H.K. Genant, J. Musculoskelet. Neuronal Interact. 1
(2000) 45–51.
[62] C. Schmidt, M. Priemel, T. Kohler, A. Weusten, R. Muller, M. Amling, F.
Eckstein, J. Bone Miner. Res. 18 (2003) 1486–1496.
[63] C. Hanusch, S. Hoeger, G.C. Beck, Methods 43 (2007) 68–78.
[64] J.C. Marini, B. Lee, P.J. Garlick, Lab. Anim. 40 (2006) 275–281.
[65] M. Dambacher, M. Ito, R. Müller, M. Neff, L. Qin, E. Schacht, S. Schmitt, Y.L.
Zhao, J. Miner. stoffwechs. 11 (2004) 13–19.
[66] V. Patel,A.S. Issever, A. Burghardt, A. Laib, M. Ries, S. Majumdar, J. Orthop. Res.
21 (2003) 6–13.
[67] H. Watanabe, E.H. Kislauskis, C.A. Mackay, A. Mason-Savas, S.C. Marks Jr., J.
Cell Sci. 111 (Pt. 9) (1998) 1287–1292.
[68] C. Wietholt, D.L. Roerig, J.B. Gordon, S.T. Haworth, R.C. Molthen, A.V. Clough,
Eur. J. Nucl. Med. Mol. Imaging 35 (2008) 1124–1132.
[69] M. Drinane, J. Mollmark, L. Zagorchev, K. Moodie, B. Sun, A. Hall, S. Shipman,
P. Morganelli, M. Simons, M.J. Mulligan-Kehoe, Circ. Res. 104 (2009)337–345.
[70] A.C. Langheinrich, D.G. Sedding, M. Kampschulte, R. Moritz, J. Wilhelm, W.G.
Haberbosch, E.L. Ritman, R.M. Bohle, Atherosclerosis 202 (2009) 103–110.
[71] N. Kleinstreuer, T. David, M.J. Plank, Z. Endre, Am. J. Physiol. Renal Physiol.
294 (2008) F1453–F1464.
[72] A. Garcia-Sanz, A. Rodriguez-Barbero, M.D. Bentley, E.L. Ritman, J.C. Romero,
Hypertension 31 (1998) 440–444.
[73] L.A. Fortepiani, M.C.O. Ruiz, F. Passardi, M.D. Bentley, J. Garcia-Estan,
E.L. Ritman, J.C. Romero, Am. J. Physiol. Renal Physiol. 285 (2003)
F852–F860.
[74] M.D. Bentley, S.M. Jorgensen, L.O. Lerman, E.L. Ritman, J.C. Romero, Anat. Rec.
(Hoboken) 290 (2007) 277–283.
[75] S. Ananda, V. Marsden, K. Vekemans, E. Korkmaz, N. Tsafnat, L. Soon, A. Jones,
F. Braet, J. Electron Microsc. 55 (2006) 151–155.
[76] S.D. Murad, V.A.L. Dom, E.L. Ritman, P.C. de Groen, P.E. Beigley, S.C. Abraham,
P.E. Zondervan, H.L.A. Janssen, J. Gastroenterol. Hepatol. 23 (2008) 1561–
1566.
[77] T. Abruzzo, L. Tumialan, C. Chaalala, S. Kim, R.E. Guldberg, A. Lin, J. Leach, J.C.Khoury, A.E. Morgan, C.M. Cawley 3rd, Synapse 62 (2008) 557–565.
[78] S.B. Clauss, D.L. Walker, M.L. Kirby, D. Schimel, C.W. Lo, Dev. Dyn. 235 (2006)
2786–2794.
[79] A.S. Brown, L. Leamen, V. Cucevic, F.S. Foster, Invest. Ophthalmol. Vis. Sci. 46
(2005) 2231–2237.
[80] A. Dorr, J.G. Sled, N. Kabani, Neuroimage 35 (2007) 1409–1423.
[81] M. Gossl, P.E. Beighley, N.M. Malyar, E.L. Ritman, Am. J. Physiol. Heart Circ.
Physiol. 287 (2004) H2346–H2351.
[82] K.A. Johnson, Toxicol. Pathol. 35 (2007) 59–64.
[83] R.I. Lopez-Soler, M.P. Brennan, A. Goyal, Y. Wang, P. Fong, G. Tellides, A.
Sinusas, A. Dardik, C. Breuer, J. Surg. Res. 139 (2007) 1–6.
[84] C.T. Badea, B. Fubara, L.W. Hedlund, G.A. Johnson, Mol. Imaging 4 (2005)110–
116.
[85] C.T. Badea, L.W. Hedlund, M. Lin, J. Mackel, G.A. Johnson, Contrast Media Mol.
Imaging 1 (2006) 153–164.
[86] A. Chouker, M. Lizak, D. Schimel, T. Helmberger, J.M. Ward, D. Despres, I.
Kaufmann, C. Bruns, F. Lohe, A. Ohta, M.V. Sitkovsky, B. Klaunberg, M. Thiel,
Invest. Radiol. 43 (2008) 77–91.
[87] S.A. Detombe, N.L. Ford, F. Xiang, X. Lu, Q. Feng, M. Drangova, Invest. Radiol.43 (2008) 520–529.
[88] K.C. Graham, N.L. Ford, L.T. MacKenzie, C.O. Postenka, A.C. Groom, I.C.
MacDonald, D.W. Holdsworth, M. Drangova, A.F. Chambers, Invest. Radiol. 43
(2008) 92–99.
[89] T. Henning, A.W. Weber, J.S. Bauer, R. Meier, J.M. Carlsen, E.J. Sutton, S.
Prevrhal, S.I. Ziegler, H. Feussner, H.E. Daldrup-Link, E.J. Rummeny, Acad.
Radiol. 15 (2008) 342–349.
[90] A.C. Langheinrich, E.L. Ritman, Invest. Radiol. 41 (2006) 645–650.
[91] I. Willekens, T. Lahoutte, N. Buls, C. Vanhove, R. Deklerck, A. Bossuyt, J. de
Mey, Mol. Imaging Biol. 11 (2009) 128–135.
[92] V. Persy, A. Postnov, E. Neven, G. Dams, M. De Broe, P. D’Haese, N. De Clerck,
Arterioscler. Thromb. Vasc. Biol. 26 (2006) 2110–2116.
[93] E. Namati, D. Chon, J. Thiesse, E.A. Hoffman, J. de Ryk, A. Ross, G. McLennan,
Phys. Med. Biol. 51 (2006) 6061–6075.
[94] S.H. Bartling, J. Dinkel, W. Stiller, M. Grasruck, I. Madisch, H.U. Kauczor, W.
Semmler, R. Gupta, F. Kiessling, Eur. Radiol. 18 (2008) 1375–1384.
[95] X.F. Li, P. Zanzonico, C.C. Ling, J. O’Donoghue, Technol. Cancer Res. Treat. 5
(2006) 147–155.
12 S.J. Schambach et al./ Methods 50 (2010) 2–13
8/10/2019 Schambach 2010
http://slidepdf.com/reader/full/schambach-2010 12/12
[96] Y. Hori, N. Takasuka, M. Mutoh, T. Kitahashi, S. Kojima, K. Imaida, M. Suzuki,
K. Kohara, S. Yamamoto, N. Moriyama, T. Sugimura, K. Wakabayashi, Cancer
Sci. 99 (2008) 1774–1777.
[97] S. Shofer, C. Badea, Y. Qi, E. Potts, W.M. Foster, G.A. Johnson, J. Appl. Physiol.
105 (2008) 669–677.
[98] A.A. Postnov, K. Meurrens, H. Weiler, D. Van Dyck, H. Xu, P. Terpstra, N.M. De
Clerck, J. Microsc. 220 (2005) 70–75.
[99] M.F. Fredin, L. Hultin, G. Hyberg, E. Rehnstrom, E. Hultgren Hornquist, S.
Melgar, L. Jansson, Inflamm. Bowel Dis. 14 (2008) 491–499.
[100] X.Y. Zhu, M. Rodriguez-Porcel, M.D. Bentley, A.R. Chade, V. Sica, C. Napoli, N.
Caplice,E.L. Ritman,A. Lerman,L.O. Lerman,Circulation109 (2004)2109–2115.[101] M. Drangova, N.L. Ford, S.A. Detombe, A.R. Wheatley, D.W. Holdsworth,
Invest. Radiol. 42 (2007) 85–94.
[102] L.R. Desnoyers, R. Pai, R.E. Ferrando, K. Hotzel, T. Le, J. Ross, R. Carano, A.
D’Souza, J. Qing, I. Mohtashemi, A. Ashkenazi, D.M. French, Oncogene 27
(2008) 85–97.
[103] M. Almajdub, M. Nejjari, G. Poncet, L. Magnier, E. Chereul, C. Roche, M. Janier,
Contrast Media Mol. Imaging 2 (2007) 88–93.
[104] M. Gossl, M.D. Bentley, L.O. Lerman, Nephron Clin. Pract. 103 (2006) c66–c70.
[105] L.A. Fortepiani, M.C. Ruiz, F. Passardi, M.D. Bentley, J. Garcia-Estan, E.L.
Ritman, J.C. Romero, Am. J. Physiol. Renal Physiol. 285 (2003) F852–F860.
[106] M. Almajdub, L. Magnier, L. Juillard, M. Janier, Contrast Media Mol. Imaging 3
(2008) 120–126.
[107] B.Y. Durkee, S.R. Mudd, C.N. Roen, L. Clipson, M.A. Newton, J.P. Weichert, P.J.
Pickhardt, R.B. Halberg, Acad. Radiol. 15 (2008) 334–341.
[108] P. Choquet, A. Calon, E. Breton, F. Beck, C. Domon-Dell, J.N. Freund, A.
Constantinesco, C. R. Biol. 330 (2007) 821–827.
[109] Y.K. Luu, S. Lublinsky, E. Ozcivici, E. Capilla, J.E. Pessin, C.T. Rubin, S. Judex,
Med. Eng. Phys. 31 (2009) 34–41.
[110] B.P. Chugh, J.P. Lerch, L.X. Yu, M. Pienkowski, R.V. Harrison, R.M. Henkelman,
J.G. Sled, Neuroimage 47 (2009) 1312–1318.
[111] A. Proweller, A.C. Wright, D. Horng, L. Cheng, M.M. Lu, J.J. Lepore, W.S. Pear,
M.S. Parmacek, Proc. Natl. Acad. Sci. USA 104 (2007) 16275–16280.
[112] S. Heinzer, T. Krucker, M. Stampanoni, R. Abela, E.P. Meyer, A. Schuler, P.
Schneider, R. Muller, Neuroimage 32 (2006) 626–636.
[113] R. Mizutani, A. Takeuchi, K. Uesugi, M. Ohyama, S. Takekoshi, R.Y. Osamura, Y.
Suzuki, Brain Res. 1199 (2008) 53–61.
[114] A. de Crespigny, H. Bou-Reslan, M.C. Nishimura, H. Phillips, R.A. Carano, H.E.D’Arceuil, J. Neurosci. Methods 171 (2008) 207–213.
[115] E.W. Newcomb, S. Demaria, Y. Lukyanov, Y. Shao, T. Schnee, N. Kawashima, L.
Lan, J.K.Dewyngaert, D. Zagzag, W.H. McBride, S.C.Formenti, Clin. CancerRes.
12 (2006) 4730–4737.
[116] T. Engelhorn, I.Y. Eyupoglu, M.A. Schwarz, M. Karolczak, H. Bruenner, T.
Struffert, W. Kalender, A. Doerfler, Neurosci. Lett. 458 (2009) 28–31.
[117] D. Balvay, I. Tropres, R. Billet, A. Joubert, M. Peoc’h, C.A. Cuenod, G. Le Duc,
Radiology 250 (2009) 692–702.
[118] J.M. Yuhas, Radiat. Res. 60 (1974) 321–332.
[119] D. Bhattacharjee, A. Ito, In Vivo 15 (2001) 87–92.
[120] M. Takahashi, S. Kojima, K. Yamaoka, E. Niki, Radiat. Res. 154 (2000) 680–
685.
[121] W. Stiller, M. Kobayashi, K. Koike, U. Stampfl, G.M. Richter, W. Semmler, F.
Kiessling, Rofo 179 (2007) 669–675.
[122] L.A. Feldkamp, L.C. Davis, J.W. Kress, J. Opt. Soc. Am. 1 (1984) 612–619.
S.J. Schambach et al./ Methods 50 (2010) 2–13 13