Download In-laboratory diffraction-enhanced x-ray imaging (DEXI

In-laboratory diffraction-enhanced x-ray imaging (DEXI) of an equine hoof
Daniel P. Fogarty Ph.D.1†, Benjamin Reinhart B.S.1, Tochko Tzvetkov Ph.D.1, Ivan Nesch J.D.,
Ph.D.1, and Cooper Williams V.M.D.2
Nesch, LLC 9800 Connecticut Drive, Crown Point, IN 46307.
Cooper Williams, V.M.D. 720 Houcksville Rd, Hampstead, MD 21074.
† Author to whom correspondence should be addressed. Electronic mail: [email protected].
ABSTRACT
We describe and demonstrate the first application of a laboratory-based diffraction-enhanced xray imaging (DEXI) instrument for non-invasive equine imaging.
A formalin preserved
disarticulated forelimb from a near-term aborted miniature horse fetus is imaged with DEXI. The
resultant calculated images—absorption, extinction, refraction, and scattering—are presented,
and soft-tissues such as the dorsal digital extensor tendon, articular cartilage, as well as various
joint, tendon sheath, and bursa recesses are observed in simultaneous registration with the
adjacent dense bone tissue. Radiation dose calculations were performed and a calculated surface
dose of 0.6 mGy for the soft muscular tissue was determined for the imaging experiment.
KEY WORDS: Diffraction-Enhanced X-ray Imaging; Analyzer-based imaging; Lameness and
Imaging; Radiology; Diagnostic Imaging.
I. INTRODUCTION
X-ray radiography has played a significant role in the diagnosis and treatment of multiple types
of pathological and systemic medical diseases and conditions.
The primary advantage of
conventional radiography is the ability to visualize the internal structure of a living species noninvasively; dense materials or high Z elements located within an object or specimen, such as
bone matter can be visualized quite easily using this technique. Soft-tissues, however, are
difficult to detect using conventional radiography because of the tissue‟s low absorptivity and
resultant low absorption contrast between neighboring soft-tissue structures.
As a result,
pathological conditions that reside in regions composed of soft tissue must be diagnosed using
alternative, non-invasive techniques.
To date these techniques have included magnetic
resonance imaging (MRI) and ultrasonography—both of which have limitations.
A new class of x-ray imaging techniques, analyzer-based imaging, has emerged from research
performed at synchrotron radiation facilities [1-6]. Analyzer-based imaging involves placing an
analyzer crystal between the object of interest and the x-ray detector or film. By placing the
analyzer crystal between the sample and the x-ray detector, one can produce high contrast
images that are normally not associated with x-ray imaging.
Recently, we have designed and demonstrated the first low-energy diffraction-enhanced x-ray
imaging (DEXI) instrument which utilizes a conventional x-ray tube as its source of x-rays [7].
In the work presented here, we demonstrate the first application of DEXI technology for noninvasive equine imaging. A formalin preserved disarticulated forelimb from a near-term aborted
miniature horse fetus has been imaged using the DEXI instrument, and the resultant computed
1
images show soft-tissue structures that are not visible using conventional radiography. These
preliminary images show great promise for the use of this technology to assist in the detection,
diagnosis, and potentially the prevention of harmful veterinary related conditions and injuries.
II. MATERIALS & METHODS
A. Formalin preserved equine hoof
Imaging was performed on a disarticulated forelimb from a near-term aborted miniature horse
fetus. The sample was removed just above the metacarpophalangeal joint and preserved in
formalin.
B. DEXI Instrument
A detailed description of the laboratory-based DEXI instrument used in this work can be found
in a previous publication [7]. In summary, the DEXI instrument utilizes a 2.2-kW silver x-ray
source, which is operated at a voltage of 55 kV and a current of 35 mA. A mismatched, twocrystal monochromator is placed in front of the conventional x-ray source and the resulting
monoenergetic 22.16-keV beam measuring 12 mm × 60 mm is passed through the sample. An
analyzer crystal is placed immediately after the sample, followed by a high-resolution digital
detector. The DEXI instrument has an experimentally determined spatial resolution of 160±7
μm in the horizontal direction and 153±7 μm in the vertical direction [7].
C. Imaging procedure
2
The general DEXI operating protocol requires that the user selects three parameters: (1) number
of angular positions for the analyzer crystal [1]; (2) the exposure time per angular position, and
(3) number of sample positions (translational and/or rotational) for the object.
Depending upon the mode of operation (e.g. DEI, MIR, etc.), the number of analyzer-crystal
angular positions could range from a single position to multiple positions [1,8]. Once the
angular positions and the sample‟s translational positions are determined, the user will then
choose the desired exposure time per angular position and then start the experiment.
After the image acquisition is complete the user can choose from a number of image processing
methods. For the data presented here, multiple-imaging radiography (MIR) methodology was
used to compute the reported images [8-10]. MIR, involves collecting multiple raw images of
the specimen by placing the analyzer crystal at multiple angular positions along the rocking
curve. A set of four new images are computed using these raw data collected at each analyzer
position. The resultant computed images are apparent absorption, extinction, refraction, and
scattering images.
D. Radiation dose calculations
To calculate the radiation dose deposited for this experiment we have used the following
approach:
The absorbed dose for x-rays can be described by
(1)
3
In the DEXI instrument we employ a monochromatic x-ray beam of 22.16 keV; as a result, Eqn.
1 becomes
(2)
where Nabs is the number of photons that result in energy absorption, Eph is the photon energy
and M is the mass of the material. The number of photons absorbed that result in energy
absorption can be expressed as:
(3)
The μ/ρ values (mass attenuation coefficient) are specific to a material and are independent of the
state (i.e. solid, gas, liquid). They are directly related to the cross section per atom. The total
mass attenuation coefficient, μT/ρ, accounts for all x-ray loss mechanisms—photoelectric
absorption, elastic scattering and Compton scattering, and is used to calculate the number of xrays transmitted through a material. The energy absorption mass attenuation coefficient, μEN/ρ, is
a measure of those x-rays that are absorbed and result in energy deposition (all of the
photoelectric absorption and part of the Compton scattering).
Substituting Eqn. 3 into Eqn. 2 gives the following:
(4)
where A is the cross sectional area of the region the beam hits and t is the thickness of the mass
of density ρ in which the dose is to be calculated.
4
In the x-ray imaging regime, the maximum dose occurs at the surface of objects. Thus the
largest dose is the “surface dose”. This is the dose that is usually quoted when talking about
dose to objects and tissues in imaging. The surface dose is the zero thickness limit of Eqn. 4
(5)
Therefore, the surface dose is the energy deposited per unit area times the energy absorption
mass attenuation coefficient. Eqn. 5 is used to evaluate the surface dose accumulated by the
equine hoof tissues for both muscle type tissues and bone type tissues. The values for μEN/ρ for
muscle type tissues (Muscle Skeletal ICRU-44; 0.415 cm2/g) and bone type tissues (Bone
Cortical; 2.75 cm2/g) are taken from the National Institute of Standards and Technology (NIST)
[11].
III. RESULTS
A. DEXI images
A DEXI data set of the formalin preserved equine hoof is presented in Fig. 1. The specimen
measures approximately 8 cm × 3.5 cm × 3 cm and was imaged in air under ambient conditions.
Images of the specimen were taken using 17 analyzer positions with an exposure time of 300
seconds per angular position. The specimen was translated horizontally 4 times due to it being
wider than the 12 × 60 mm2 x-ray beam field of view.
5
It is important to note that the aim of the work presented here is not to provide new anatomical
findings of an equine appendage, but rather to illustrate the capability of the DEXI instrument to
image these types of specimens.
The calculated absorption image presented in Fig. 1A is representative of an absorption image
taken by a conventional clinical x-ray source. The primary structures that are observed in the
absorption image are dense bone structures of the equine hoof— specifically the first, second,
and third phalanges and the distal sesamoid (navicular) bone. Tendons and other soft tissues are
not observed in this type of contrast image.
Details of the surrounding soft-tissue structures are observed in Figs. 1B-D.
Soft-tissue
structures such as the dorsal digital extensor tendon, articular cartilage, as well as various joint,
tendon sheath, and bursa recesses are observed in these images; the soft-tissue structures of the
equine hoof are best seen in the refraction image (Fig. 1C). Fig. 2 presents a close up of the
same refraction image shown in Fig. 1C. The soft-tissue structures of the forelimb are labeled
and a list of the corresponding anatomical structures is provided in the figure caption.
B. Calculated radiation dose
The formalin preserved equine hoof was imaged using 17 analyzer positions with an exposure
time of 300 seconds per angular position. Using Eqn. 5 described in Section IID and an
experimentally measured incident surface fluence rate of Φ = 812 mm-2s-1, the calculated surface
dose acquired during this imaging procedure is 0.6 mGy.
6
Assuming the thickness of the
surrounding soft tissue is negligible, the calculated upper limit of the surface dose for the dense
bone structures of the equine hoof is 4 mGy.
IV. DISCUSSION
The soft-tissues that are observed in Figs. 1 and 2 are the result of the analyzer crystal‟s
sensitivity to absorption, refraction and small-angle scattering occurring within the sample.
Materials (or in this case soft tissues) that have either a difference in their absorption, refraction,
and/or scattering properties can be distinguished unlike in a conventional radiograph.
Additionally, those structures that have identical or similar chemical properties (i.e. identical
absorptivities) can be detected in a refraction image due to the interface(s) found between
neighboring structures as observed in the refraction image presented in Fig. 1C.
In addition to the detection of soft-tissue structures, it is important to note that all DEXI
images—extinction, refraction, and scattering—are inherently in registry with one another. Each
corresponding pixel of the three types of images, represent the same location within the object.
As a result, this type of overlap provides a unique perspective for determining the relative
positions of the internal structures of a sample.
The estimated surface dose for both the soft tissues and hard tissues were calculated to be 0.6
mGy and 4 mGy respectively. For the soft-tissue structures, the calculated surface dose is much
less than the radiation dose acquired for the same type of tissue during a conventional
radiography scan (3-4 mGy) [12]. The calculated surface dose for the hard-tissue structures,
however, is equivalent as a result of the dense bone material and the x-ray energy of the
7
conventional radiation source used in the DEXI instrument. The 22-keV monochromatic x-ray
beam used in this experiment is not normally used when imaging dense bone material—typically
higher energy x-rays are used. The bone material absorbs significantly more radiation at this
energy compared to the surrounding soft-tissue as was shown by the calculations; this
phenomenon is further illustrated by the observed imaging artifacts, the „salt and pepper‟ noise,
found in both the refraction and scattering images presented in Figs. 2C and 2D.
The salt and pepper noise observed in each of the images presented in Fig. 2 in regions where
hard tissues, such as the first and second phalanges, are located is the result of the low number of
photons that reach the x-ray detector during raw image collection. The low number of photons
reaching the detector results in a low signal to noise ratio; as a result, areas in which there are
dense, bone structures result in randomized black and white pixilated regions. This type of
imaging artifact can be overcome by either increasing the imaging time or using a more powerful
x-ray source.
A high-energy DEXI instrument is currently being actively pursued. The current low-energy
DEXI instrument is capable of imaging smaller, less-dense materials or specimens, such as small
animals; the equine hoof presented in this work is near the upper limit of what the current DEXI
machine can reasonably image. A high-energy DEXI instrument will provide a much greater
photon flux that will allow larger more dense objects to be imaged, such as an adult equine hoof.
The high-energy x-ray source will significantly reduce the radiation dose incident on the
specimen and greatly reduce imaging time.
8
We have presented the first set of diffraction-enhanced x-ray images of an equine hoof taken on
a laboratory-based DEXI instrument. The DEXI images were produced using the multipleimaging radiography computing methodology; these data reveal anatomical structures that are
normally undetectable using conventional x-ray techniques while using far less radiation
exposure to obtain the high contrast images. These results give great promise for the use of
DEXI for non-invasive equine imaging and other veterinary related fields.
V. ACKNOWLEDGEMENTS
The authors would like to thank the Indiana 21st Century Research and Technology fund and the
Indiana Economic Development Corporation for providing the financial support for this work.
The authors would also like to thank Professor Dean Chapman for his guidance on the radiation
dose calculations and his helpful discussions and feedback.
VI. REFERENCES
[1] Chapman D, Thomlinson W, Johnston RE, et al. Diffraction enhanced x-ray imaging. Phys.
Med. Biol. 1997; 42: 2015-25.
[2] Davis TJ, Gao D, Gureyev TE, Stevenson AW, Wilkins SW. Phase-contrast imaging of
weakly absorbing materials using hard X-rays. Nature. 1995; 373: 595-8.
[3] Lewis R. Medical applications of synchrotron radiation x-rays. Phys. Med. Biol. 1997; 42:
1213-43.
[4] Coan P, Mollenhauer J, Wagner A, Muehleman C, Bravin A. Analyzer-based imaging
technique in tomography of cartilage and metal implants: A study at the ESRF. Eur. J. Radiol.
2008; 685: S41-8.
9
[5] Bech M, Jensen TH, Feidenhans R, et al. Soft-tissue phase-contrast tomography with an xray tube source. Phys. Med. Biol. 2009; 54: 2742-53
[6] Parham C, Zhong Z, Connor DM, Chapman LD, Pisano ED. Design and implementation of
a compact low-dose diffraction enhanced medical imaging system. Acad. Radiol. 2009; 18:
911-7.
[7] Nesch I, Fogarty DP, Tzvetkov T, et al. The design and application of an in-laboratory
diffraction-enhanced x-ray imaging instrument. Rev. Sci. Instrum. 2009; 80: 093702.
[8] Wernick MN, Wirjadi O, Chapman D, et al. Multiple-image radiography. Phys. Med. Biol.
2003; 48: 3875-95.
[9] Chou CY, Anastasio MA, Brankov JG, et al. An extended diffraction-enhanced imaging
method for implementing multiple-image radiography. Phys. Med. Biol. 2007; 52: 1923-45.
[10] Khelashvili G, Brankov JG, Chapman D, et al. A physical model of multiple-image
radiography. Phys. Med. Biol. 2006; 51: 221-36.
[11] ICRU (1989), Tissue Substitutes in Radiation Dosimetry and Measurement, Report 44 of
the International Commission on Radiation Units and Measurements (Bethesda, MD).
[12] Brenner DJ, Hall EJ. Computed Tomography—An Increasing Source of Radiation
Exposure. N. Engl. J. Med. 2007; 357: 2277-84.
FIGURE CAPTIONS:
Fig. 1. A DEXI data set of a formalin preserved equine hoof: (A) calculated absorption image,
(B) extinction image, (C) refraction image, and (D) scattering image. Observed anatomical hard10
tissue structures are noted: (P1) first phalanx, (P2) second phalanx, (P3) third phalanx, and
(DSB) the distal sesamoid (navicular) bone.
Fig. 2. Close up of the refraction image presented in Fig. 1c. Observed anatomical features: (A)
deep digital extensor tendon; (B) straight sesamoidean ligament; (C) middle scutum; (D) digital
tendon sheath; (E) toric part digital cushion; (F) collateral sesamoidean ligament; (G)
podotrochlear bursa; (H) cuneal part digital cushion; (I) distal impar ligament; (J) articular
cartilage DIP joint; (K) laminae of the corium parietis; (L) stratum internum; (M) stratum
medium; (N) stratum externum; (O) pulvinus coronae; (P) insertion dorsal digital extensor
tendon; (Q) dorsal recess DIP joint; (R) articular cartilage PIP joint; (S) dorsal recess PIP joint;
(T) corium soleae; and (U) corium cunei.
11