Download ARTICLE IN PRESS Cadaveric and human

ZEMEDI-10433; No. of Pages 6
ARTICLE IN PRESS
ORIGINALARBEIT
Cadaveric and in vivo human joint imaging based on differential
phase contrast by X-ray Talbot-Lau interferometry
Junji Tanaka 1,∗ , Masabumi Nagashima 2 , Kazuhiro Kido 3 , Yoshihide Hoshino 3 , Junko Kiyohara 3 ,
Chiho Makifuchi 3 , Satoshi Nishino 3 , Sumiya Nagatsuka 3 , Atsushi Momose 4
1
Department of Radiology, Saitama Medical University, 38 Morohongo, Moroyama, Iruma, Saitama 350-0495, Japan
Department of Anatomy, Saitama Medical University, 38 Morohongo, Moroyama, Iruma, Saitama 350-0495, Japan
3
KonicaMinolta Medical and Graphic, Inc., 2970 Ishikawa-machi, Hachioji, Tokyo 192-8505, Japan
4
Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, 2-1-1 Katahira, Aoba-ku, Sendai, Miyagi
980-8577, Japan
2
Received 4 September 2012; accepted 9 November 2012
Abstract
We developed an X-ray phase imaging system based on
Talbot-Lau interferometry and studied its feasibility for
clinical diagnoses of joint diseases. The system consists of
three X-ray gratings, a conventional X-ray tube, an object
holder, an X-ray image sensor, and a computer for image
processing. The joints of human cadavers and healthy volunteers were imaged, and the results indicated sufficient
sensitivity to cartilage, suggesting medical significance.
Keywords: X-ray interferometry, a new imaging
technology, in vivo imaging, cadavers and normal
volunteers
1 Introduction
The principle of medical X-ray imaging has not changed
since the discovery of X-rays in 1895, in which image contrast depends on the attenuation of X-rays. Phase shifts also
Darstellung von menschlichen Gelenken,
in vivo und bei Verstorbenen, basierend auf
Differential-Phasenkontrast durch
Röntgen-Talbot-Lau-Interferometrie
Zusammenfassung
Wir entwickelten ein Röntgen-Phasen-Imaging-System,
basierend auf Talbot-Lau-Interferometrie und bestehend
aus drei Röntgengittern, einer konventionellen Röntgenröhre, einem Objekttisch, einem Röntgenbildsensor
und einem Computer zur Bildverarbeitung. Anschließend
untersuchten wir seine Nutzbarkeit für die Diagnostizierung von Gelenkerkrankungen, wobei die Gelenke von
gesunden Probanden sowie von Verstorbenen dargestellt
werden konnten. Die Ergebnisse zeigten eine ausreichende
Sensitivität für Knorpel, was für die klinische Bedeutsamkeit unseres entwickelten Systems spricht.
Schlüsselwörter: Röntgeninterferometrie, neue
Röntgentechnik, Verstorbene und gesunde Probanden
occur when X-rays pass through an object. If the phase shifts
could be captured and imaged, an extremely high sensitivity
would be attained because the interaction cross-section of the
X-ray phase shift is about 1000 times larger than that of the
attenuation, especially for light materials [1]. Thanks to the
∗ Corresponding author: Junji Tanaka, Department of Radiology, Saitama Medical University, 38 Morohongo, Moroyama, Iruma, Saitama 350-0495, Japan.
E-mail: [email protected] (J. Tanaka).
Z. Med. Phys. xxx (2012) xxx–xxx
http://dx.doi.org/10.1016/j.zemedi.2012.11.004
http://journals.elsevier.de/zemedi
ARTICLE IN PRESS
2
J. Tanaka et al. / Z. Med. Phys. xxx (2012) xxx–xxx
development of X-ray digital imaging technology, novel X-ray
imaging techniques were developed in the 1990s for visualizing low-contrast structures by quantitatively processing
X-ray phase-contrast images. This is referred to as “X-ray
phase imaging.” However, it is still not available for medical
application because of the difficulty in its implementation in
hospitals.
X-ray Talbot interferometry based on the Talbot effect [2]
is a relatively new differential phase-contrast technique, in
which the refraction of X-rays is captured through interference and is visualized as distortion of the Moiré pattern [3,4].
It is characterized by the use of gratings and is also called
“grating interferometry.” Since it can be used with broadband energy X-rays, its application in the medical setting
has been expected. However, since the X-ray Talbot interferometer requires partial spatial coherence, the X-ray source
available is limited to either a microfocus X-ray tube or a
synchrotron. Since the former cannot emit X-rays of sufficient flux and the latter is enormous, the clinical application
of phase imaging has not been feasible. The use of an X-ray
source of reasonable size and sufficient power is needed.
F. Pfeiffer et al. solved this problem by X-ray TalbotLau interferometry [5], where a conventional X-ray tube is
employed with a source grating in addition to the configuration of the Talbot interferometer. X-ray phase imaging can be
performed with a reasonable exposure time, and we studied its
feasibility for the diagnoses of joint diseases. We constructed
an X-ray Talbot-Lau interferometer based on a design with a
wave-optic simulation [6], and used it to capture images of
joints of cadavers and healthy volunteers. We report here the
first imaging results and discuss their clinical significance.
2 Material and methods
2.1 System configuration
When a grating (hereafter, G1) is illuminated by X-rays
of partial spatial coherence, self-images appear periodically
downstream of G1 as a result of the Talbot effect. When the
self-image is overlaid with another grating (G2) that has the
same pitch as that of a self-image, a Moiré pattern appears. If
an object is placed in front of G1, the wave-front of the X-rays
is distorted and the Moiré pattern is also distorted. This is the
Talbot interferometer. The phase information of the object can
be extracted from the distorted Moiré patterns by the use of
the fringe-scanning method [3]. In order to use an incoherent
X-ray tube, a source grating (G0) is added near the tube, as
shown in Fig. 1, to configure the Talbot-Lau interferometer
[5]
With Talbot-Lau interferometry, three different types of
images can be simultaneously obtained by calculations: an
attenuation image, a small-angle-scattering (dark-field) image
[7], and a differential phase image. When X-rays pass an
object, some X-rays are absorbed, and other X-rays are
refracted by the object. The former effect results in an
Figure 1. Schematic illustration of a Talbot interferometer modified
with the Lau effect. Self-images 1, 2 and 3 are due to the X-ray beams
B1, B2 and B3, respectively, from the slits of G0. The beams from
G0 produce self-images constructively at the G2 plane, allowing the
use of a conventional X-ray source.
attenuation image, and the latter results in a differential
phase image. When X-rays are refracted strongly, a corresponding contrast appears in the small-angle-scattering image
[8]. For clinical diagnoses, the three images can be used
complementarily.
The basic design and specifications of our system were
based on a simulation by W. Yashiro et al. [6]. An X-ray generator of a tungsten anode was operated with a tube voltage
of 40 kVp, which was selected according to the simulation,
and a tube current of 75 mA. The mean X-ray energy was 28
keV. The pitches of G0, G1 and G2 were 22.8 ␮m, 4.3 ␮m,
and 5.3 ␮m, respectively. The opening width of G0 was 7 ␮m,
and the duty cycle of G1 and G2 was 0.5. G1 and G2 were
located 1.1 m and 1.36 m from G0, respectively. G1 was a ␲/2
phase grating for 28 keV X-rays, and G2 was an amplitude
grating, whose gold pattern height was 43 ␮m. The object to
be imaged was placed in front of G1. The area size of G1 and
G2 was 60 mm x 60 mm, and the effective field of view was
49 mm x 49 mm taking account of the magnification of the
image. All of the gratings were fabricated by X-ray lithography and gold electroplating [9]. A flat panel detector was
located behind G2 and its pixel size was 85 ␮m. A picture of
the system is shown in Fig. 2(a).
The FWHM values of the focal spot size of the X-ray
tube and the LSF of the detector were 350 ␮m and 90 ␮m,
respectively. The system spatial resolution was calculated to
be 99 ␮m from the FWHM values [6]. For the measurement
of a differential phase image, a five-step fringe-scan was performed.
ARTICLE IN PRESS
J. Tanaka et al. / Z. Med. Phys. xxx (2012) xxx–xxx
3
Figure 2. Overviews of our Talbot-Lau interferometers: (a) A system for cadaveric specimens, where X-rays pass from the left to the right,
through G0, the target object, G1, and G2 to the X-ray detector, and (b) A system for in vivo imaging, where X-rays pass from the upside
to the downside.
A system for in vivo imaging was next constructed with a
vertical configuration, as shown in Fig. 2(b). The specifications of the system are the same as those of Fig. 2(a) except
for the X-ray tube. The focal spot size of the X-ray tube was
changed to 450 ␮m (FWHM) to shorten the exposure time,
and the system spatial resolution was 113 ␮m. The X-ray tube
was operated with a tube voltage of 40 kVp and a tube current of 100 mA. A three-step fringe-scan was performed with
this system, and the total scan time for obtaining a differential
phase image was 32 seconds including 19 seconds of X-ray
exposure.
2.2 Human cadaveric specimens
We first imaged small parts of a cadaver, such as a finger,
with the approval of the ethics committee of Saitama Medical University, Japan. A cadaveric hand was procured from
the anatomy laboratory of Saitama Medical University (it had
been preserved in 70% ethanol for one year after immersion
fixation with 10% formalin for two weeks after the death of
the donor).
Figure 3 shows the first images of a cadaveric thumb
obtained by the system shown in Fig. 2(a). When these
images were measured, the grating lines were oriented so that
differential features were sensed in the horizontal direction.
The fingers were oriented horizontally in Fig. 3(c). For each
image-acquisition, the average radiation skin dose was 9
mGy, which was measured using an ionization-chamber
dosimeter. The resultant images were evaluated and correlated
with macroscopic anatomical structures by a radiologist (J.T.)
and an anatomist (M.N.).
2.3 Tests on healthy volunteers
With the approval of the ethics committee, in vivo imaging tests on 10 healthy volunteers followed the tests on the
cadaveric specimens. The images were obtained with the system shown in Fig. 2(b). A volunteer sat by the machine and
placed his/her hand on the imaging table, underneath which
G1 was located. The hand was aligned and held in the field of
view, which was indicated by a guide light. The direction of
the fingers was perpendicular to the grating lines so that the
structures in joints could be effectively visualized. Figure 4
shows one of the images of the metacarpophalangeal joint
of the third finger. Other images were similar to Fig. 4. The
average radiation skin dose was 5 mGy.
3 Results
The differential phase images showed a somewhat relieflike appearance, in which compact bone surfaces and
trabeculae of the cancellous part that run vertically are
enhanced as shown in Fig. 3. When the finger was aligned
upright, the attenuation image and the differential image
shown in Fig. 3(a) and (b) were obtained, respectively, consistently with the macroanatomic structures shown in Fig. 3(c).
The tendons such as the flexor pollicis longus, of the flexor
pollicis brevis and of the extensor pollicis were depicted
in the differential phase image, which was not shown in
the attenuation image, as Stutman D et al. reported [10].
ARTICLE IN PRESS
4
J. Tanaka et al. / Z. Med. Phys. xxx (2012) xxx–xxx
Figure 3. Images of human cadavers: (a) Attenuation image of a right thumb aligned in the upright position, and (b) corresponding differential
phase image. The tendons, such as the tendon of the flexor pollicis longus muscle (black arrows) and the tendon of the extensor pollicis muscle
(white arrows), are clearly depicted. (c) Anatomical confirmation of the right thumb. (d) Attenuation image of a metacarpophalangeal joints
of the second and third fingers, and (e) corresponding differential phase image. The surface of the cartilage is depicted, as indicated by arrows.
ARTICLE IN PRESS
J. Tanaka et al. / Z. Med. Phys. xxx (2012) xxx–xxx
5
specimens and 10 healthy volunteers. This new imaging technology has potential for clinical application such as detecting
lesions on the surface of cartilage with a resolution of around
100 ␮m. We have already proceeded to the next step with
patients.
These phase images shown here are totally distinct from
absorption radiography, therefore we need to establish protocols for diagnoses. Then, detailed analysis of the imaging
quality from a clinical point of view should be performed
extensively.
Although we consider that the current field of view is
acceptable for the first clinical use of this imaging system,
we are also enlarging the field of view to depict various other
parts of the human body.
5 Conclusion
We conducted a preclinical study of X-ray phase imaging
based on X-ray Talbot-Lau interferometry. We constructed
a system with an X-ray Talbot-Lau interferometer using a
conventional X-ray tube and installed it in a hospital. The
surface of the cartilage in finger joints was clearly depicted
with both cadavers and 10 healthy volunteers. The potential
of the system for the diagnoses of joint diseases has been
demonstrated.
Acknowledgement
Figure 4. Images of a healthy volunteer: (a) Attenuation image of a
metacarpophalangeal joint, and (b) corresponding differential phase
image. The surface of the cartilage is depicted, as indicated by arrows.
When metacarpophalangeal joints of fingers were placed in
the field of view, the attenuation image and the differential
image were obtained, as shown in Fig. 3(d) and (e) respectively. The surface of the cartilage is clearly depicted in
Fig. 3(e).
The metacarpophalangeal joint of the third finger of a
healthy volunteer is shown in Fig. 4(a) and (b), where cartilage is as clearly depicted as in Fig. 3(e). This result shows
that the fixation in formalin played no role in the cartilage
contrast and sufficient sensitivity to cartilage can be expected
in in vivo imaging.
4 Discussion
In this trial, the technique of differential phase imaging was preliminarily tested to evaluate its potential for the
diagnoses of joint diseases. The results show superior characteristics in the images of joint cartilages both of cadaveric
We express deep gratitude to the radiologic technologists: Mr. Yukihito Wada, Mr. Hiroaki Kawasaki, Mr. Masaya
Hirano, Mr. Masato Endo, Department of Radiology in
Saitama Medical University, and Mr. Felix Thiele, medical student from the Charité Universitätsmedizin Berlin, for
their advice and contributions to this study. This study was
supported by the SENTAN project of Japan Science and Technology Agency.
References
[1] Momose A. Recent Advances in X-ray Phase Imaging. Jpn J Appl Phys
2005;44:6355–67.
[2] Talbot HF. Facts Relating to Optical Science. Philos Mag 1836;9:401–7.
[3] Momose A, Kawamoto S, Koyama I, Hamaishi Y, Takai K, Suzuki
Y. Demonstration of X-ray Talbot Interferometry. Jpn J Appl Phys
2003;42:L866–8.
[4] Weitkamp T, Diaz A, David C, Pfeiffer F, Stampanoni M, Cloetens P,
et al. X-ray Phase Imaging with a Grating Interferometer. Opt Express
2005;13:6296–304.
[5] Pfeiffer F, Weitkamp T, Bunk O, David C. Phase Retrieval and Differential Phase-Contrast Imaging with Low-Brilliance X-ray Sources. Nat
Phys 2006;2:258–61.
[6] Yashiro W, Takeda Y, Momose A. Efficiency of Capturing a Phase
Image Using Cone-Beam X-ray Talbot Interferometry. J Opt Soc Am A
2008;25:2025–39.
[7] Pfeiffer F, Bech M, Bunk O, Kraft P, Eikenberry EF, Brönnimann CH,
et al. Hard X-ray Dark-Field Imaging Using a Grating Interferometer.
Nat Mat 2008;7:134–7.
ARTICLE IN PRESS
6
J. Tanaka et al. / Z. Med. Phys. xxx (2012) xxx–xxx
[8] Yashiro W, Terui Y, Kawabata K, Momose A. On the Origin
of Visibility Contrast in X-ray Talbot Interferometry. Opt Express
2010;18:16890–901.
[9] Noda D, Tanaka M, Shimada K, Hattori T. Fabrication of Diffracting Grating with Height Aspect Ratio Using X-ray Lithography
Technique for X-ray Phase Imaging. Jpn J Appl Phys 2007;46:
849–51.
[10] Stutman D, Beck TJ, Carrino JA, Bingham CO. Talbot phase-contrast
X-ray imaging for the small joints of the hand. Phys Med Biol
2011;56:5697–720.
Available online at www.sciencedirect.com