JOURNAL TRANSCRIPT
Osteoarthritis and Cartilage (2002) 10, 163–171 © 2002 OsteoArthritis Research Society International doi:10.1053/joca.2001.0496, available online at http://www.idealibrary.com on
1063–4584/02/030163+09 $35.00/0
Diffraction-enhanced X-ray imaging of articular cartilage J. Mollenhauer*†, M. E. Aurich*†, Z. Zhong‡, C. Muehleman*§, A. A. Cole*§, M. Hasnah\, O. Oltulu\, K. E. Kuettner*, A. Margulis* and L. D. Chapman\ Departments of *Biochemistry and §Anatomy, Rush Medical College, 1653 W. Congress Parkway, Chicago, Illinois 60612, U.S.A. †Department of Orthopaedic Surgery, Friedrich-Schiller-University Jena, Waldkrankenhaus ‘Rudolf Elle’, Klosterlausnitzer Strasse 81, 07607 Eisenberg, Germany ‡National Synchrotron Light Source, Brookhaven National Laboratory, Upton, New York 11973, U.S.A. \Biological, Chemical and Physical Sciences Department, Illinois Institute of Technology, 3101 South Dearborn, Chicago, Illinois 60616, U.S.A. Summary Objective: To introduce a novel X-ray technology, diffraction-enhanced X-ray imaging (DEI), in its early stages of development, for the imaging of articular cartilage. Design: Disarticulated and/or intact human knee and talocrural joints displaying both undegenerated and degenerated articular cartilage were imaged with DEI. A series of three silicon crystals were used to produce a highly collimated monochromatic X-ray beam to achieve scatter-rejection at the microradian level. The third crystal (analyser) was set at different angles resulting in images displaying different characteristics. Once the diffraction enhanced (DE) images were obtained, they were compared to gross and histological examination. Results: Articular cartilage in both disarticulated and intact joints could be visualized through DEI. For each specimen, DE images were reflective of their gross and histological appearance. For each different angle of the analyser crystal, there was a slight difference in appearance in the specimen image, with certain characteristics changing in their contrast intensity as the analyser angle changed. Conclusions: DEI is capable of imaging articular cartilage in disarticulated, as well as in intact joints. Gross cartilage defects, even at early stages of development, can be visualized due to a combination of high spatial resolution and detection of X-ray refraction, extinction and absorption patterns. Furthermore, DE images displaying contrast heterogeneities indicative of cartilage degeneration correspond to the degeneration detected by gross and histological examination. © 2002 OsteoArthritis Research Society International Key words: Articular cartilage, Osteoarthritis, X-ray, Diffraction-enhanced imaging.
tissues have little X-ray absorption contrast and, therefore, cannot be seen in a conventional radiograph. However, plain X-ray radiography has the highest resolution1, and is currently the most economical, readily available and accepted imaging technique to detect joint abnormalities, allowing the evaluation of articular cartilage only indirectly through the measurement of joint space width2–8. Consequently, conventional radiography is sensitive only in cases of advanced disease in which there has been a loss of cartilage. Focal cartilage defects or structural abnormalities in early stages of the degenerative process are generally not seen in radiographs. Non-invasive techniques, other than X-ray, that have been applied to assess the articular cartilage are based on ultrasound or magnetic resonance imaging (MRI)9–13. Imaging techniques capable of detecting and monitoring early degenerative changes in cartilage that precede the point of irreversible damage are particularly needed. Recent studies have shown that MRI is capable of delineating articular cartilage and detecting early cartilage matrix damage11–13. However, the verification and practical utility of MRI for this purpose are still being established, and there remains an interest for complementary techniques or alternatives. With such a scarcity of imaging indicators of early cartilage damage, the development of a modality that
Introduction There is a need to non-invasively detect cartilage abnormalities, especially in the initial stages of degenerative joint disease, or osteoarthritis (OA), or early in its progression. OA is a prevalent and poorly understood disease that affects the cartilage and other tissues in the joints of aging people, having a serious impact on quality of life. Information on the structure of normal cartilage and the ways in which this tissue, with limited or no ability to repair, changes after damage or disease is essential for the development of rational treatment strategies. Cartilage and other soft Received 24 January 2001; revision requested 2 May 2001; revision received 2 October 2001; accepted 9 October 2001. Supported by USAMRMC Grant DAMD 17-99-1-927 (L.C.), by US DOE DE-AC02-76CH00016 (Z.Z.) and by the State of Illinois Higher Education Cooperative Agreement (L.C.) This work was supported, in part, by NIH grants 2-P50-AR 39239 (M.A., A.M., K.K., A.C., J.M., C.M.) and GM59395-01 (M.H., O.O., L.C.), and in part by a grant from Max-Kade-Foundation (M.A.), and GlaxoWellcome, Inc. Address correspondence to: Klaus Kuettner, Ph.D. Department of Biochemistry, Rush Medical College, 1653 West Congress Parkway, Chicago, IL 60612, U.S.A. Tel.: (312)942-2129; Fax: (312)942-3053; E-mail:
[email protected]
163
164
J. Mollenhauer et al.: X-ray imaging of articular cartilage
Fig. 1. Schematic representation of the DEI setup at the synchrotron. The synchrotron beam is incident from the left and the energy is selected by the Si (3,3,3) double crystal monochromator. This beam is passed through the object and is diffracted by a matching Si (3,3,3) analyser crystal after the object. The diffracted beam is detected as an image on a phosphor image plate. The beam from the synchrotron is a fan beam that extends into and out of the plane of the paper. To create the planar image of the object, the object and the detector are scanned perpendicular to the beams as shown with the arrows.
depicts initial phases of degeneration by harnessing different tissue characteristics in cartilage than those assessed through MRI or ultrasound is warranted. In this study, we introduce diffraction-enhanced imaging (DEI), a novel X-ray technology that is in very early stages of development and biological application, for the imaging of articular cartilage. DEI is an X-ray radiographic technique that provides dramatic gains in contrast over conventional radiography by utilizing X-ray refraction and scatter rejection (extinction) as contrast mechanisms, in addition to the X-ray absorption utilized by conventional radiography14. Although DEI has the ability to acquire and compute numerous possible images from the raw data, each conveying different information concerning the properties of the tissue14,15, images used in this study were unprocessed to show the capability of DEI itself. Currently, DEI utilizes the X-ray beam delivered by a synchrotron. In principle, the DEI technique is not intrinsically tied to the synchrotron; however, another appropriately designed machine that has the level of intensity for the very short imaging times being used in the present studies does not currently exist. Continuing research by one of the authors (Chapman) is being carried out to adapt the DEI technology to X-ray instrumentation that could be utilized in a clinical setting. The details of DEI are described elsewhere14. Briefly, in DEI an imaging beam is prepared by diffracting the polychromatic beam from the synchrotron with two matching crystals to create a nearly monochromatic, highly collimated beam of a single X-ray energy (Fig. 1). This beam is then passed through the subject as in conventional radiography. In DEI, however, a third crystal (analyser crystal) is placed between the subject and the image detector. This crystal has the ability to modulate the intensity of the X-rays according to the angular deviation of X-rays through the subject, thus producing an image of high contrast and resolution. Here, we report that articular cartilage from both disarticulated and intact human knee and ankle joints can be imaged radiographically, using DEI. Furthermore, sites of contrast heterogeneities that can be identified in DE images were found to correspond to regions of cartilage degeneration identified by both gross and histological examination. It is not the intent of this study to compare DEI with other imaging techniques, but rather to present for the
first time a novel X-ray technique for the imaging of articular cartilage.
Materials and methods DEI SYSTEM
A schematic of the DEI system used for our experimentation at the Xl5A synchrotron beamline at the National Synchrotron Light Source (NSLS), Brookhaven National Laboratory is shown in Fig. 1. In this case, the collimated fan beam of X-rays is prepared by the two-crystal sets identified as the Si (3,3,3) monochromator [Si refers to the silicon crystal material and the (3,3,3) refers to the choice of the crystal orientation and lattice plane type used to diffract X-rays]. Once this beam passes through the subject, a third crystal (analyser crystal) of the same orientation and using the same reflection diffracts the X-rays onto an image-plate detector (FuJi HRV image plate, readout by a FuJi BAS2500 image-plate reader). The image of the subject is formed by scanning the subject and image plate at the same speed through the fan beam. The image plate has a pixel size of 50 m which determines the resolution of the image obtained, since the X-rays prepared by the crystal are highly collimated. The condition for X-ray diffraction (Bragg condition) from a crystal is met only when the incident beam makes the correct angle to the atomic lattice planes in the crystal for a given X-ray photon energy or wavelength. When this condition is met, the beam diffracts from the planes over a narrow range of incident angles. Thus, if the analyser crystal is rotated about an axis perpendicular to the plane shown, the crystal will go through the Bragg condition for diffraction and the diffracted intensity will trace out a profile or ‘rocking curve’16. This profile is roughly triangular and has a peak intensity close to that of the beam striking the analyser crystal. The width of this profile is typically a few microradians wide [the full width at half maximum (FWHM) is 2.1 rad at an X-ray energy of 30 keV and 3.6 rad at 18 keV, using the Si (3,3,3) reflection]. This narrow angular width provides the tools necessary to prepare and analyse the angle of X-ray beams traversing an object on the microradian scale16. Since the range of angles that can be
Osteoarthritis and Cartilage Vol. 10, No. 3 accepted by the analyser crystal is only a few microradians, the analyser crystal detects the subject’s X-ray scattering and refraction of X-rays at the microradian level (small angle scattering)17, a sensitivity which is not possible in conventional radiography. The character of the subject image, therefore, changes depending on the scattering and refraction properties of the subject. To extract refraction information, the analyser is typically set to the half intensity points on the low- and high-angle sides of the rocking curve, while the imaging takes place. For optimal extinction sensitivity, the analyser is typically set to the peak of the rocking curve during imaging. The reproducibility of the DE images is maintained by monitoring the intensity of the diffracted X-rays by the analyser just prior to imaging to ensure that the analyser is at the prescribed angular position. This intensity is monitored by a volt meter through the ion chamber. The distance between the X-ray source and the specimen is approximately 20 m while the distance between the specimen and the imaging plate is 1 m.
SPECIMENS AND IMAGING
Human tali of the ankle (talocrural) joint and an intact knee and intact ankle joint were obtained within 24 h of death of the donor through the Regional Organ Bank of Illinois with institutional IRB approval. Osteoarthritic knee cartilage and cartilage/subchondral bone pieces from knee replacement patients from the Department of Orthopedic Surgery at Rush-Presbyterian-St. Luke’s Medical Center were also obtained with donor consent and institutional IRB approval. The tali were disarticulated at imaging, but the intact knee and ankle joints were initially DE imaged in the fully articulated state complete with surrounding superficial and deep fascia—only the skin was absent. The skin could not be taken with the donor tissue due to the embalming procedure to be used on the donor remains. All specimens were categorized into levels of cartilage degeneration based on our previously published scale18,19. Tali displaying no signs of degeneration (N=4) were obtained from donors 34 to 54 years of age, and tali displaying cartilage degeneration (N=8) were obtained from donors 51 to 66 years of age. Specifically, in the case of the tali, the talar dome (the talar contribution to the ankle joint) was imaged. All specimens were imaged in a posterior to anterior direction with 18 and 30 keV X-ray energies with DEI, and by ‘conventional’ synchrotron X-ray radiography by removing the analyser crystal. The intensity of X-rays is approximately 1.7×108 photons/s/mm2 at 18 keV and 0.8×108 photons/s/mm2 at 30 keV, with the NSLS operating at 2.8 GeV with a 200 mA ring current. Early in experimentation, 10 cartilages were imaged both before and after fixation in 4% paraformaldehyde (Sigma Chemical Co, St Louis, MO, U.S.A.). After it was shown that paraformaldehyde had no detectable effect on the DE images obtained, all subsequent specimens were paraformaldehyde-fixed prior to imaging so that specimens could be collected over time. Each specimen was imaged at least twice by the same synchrotron scientist (Zhong) using identical energy and rocking curve parameters to establish reproducibility. Total exposure time for each specimen varied from 4 to 6 s. The radiation dose delivered to the specimen was measured with an ion chamber and maintained by adjusting the speed of scan, at a level of 0.3 rem for 18 keV images and 0.1 rem for 30 keV images. The X-ray radiation
165 exposure was comparable for intact and disarticulated specimens.
HISTOLOGICAL PREPARATION
After imaging, the specimens in which subchondral bone was included were decalcified in aqueous formic acid/ sodium citrate (50:50). The tissues were subsequently dehydrated in changes of ethanol at increasing concentration, paraffin embedded, sectioned to 5-m thickness and stained with picrosirius red for polarizing microscopy or Safranin O and fast green20.
Results We have applied DEI to human articular cartilages from the distal component (talus) of the ankle (talocrural) joint that were either macroscopically normal or displayed structural changes typical of early degeneration. A human ankle joint indicating the position of the talus is shown in the skeleton in Fig. 2(a). Figure 2(b) is a photo of a talar dome whose DE image, taken in the posterior to anterior position, can be seen in Fig. 2(c) at 30 keV. The articular cartilage (seen above the arrows) is clearly identifiable. A synchrotron radiograph of the same specimen, taken at the same energy without the analyser is shown in Fig. 2(d). An example of how the character of a DE image changes depending on the setting of the analyser crystal can be seen in Fig. 3. The images of the talar dome were taken with 30 keV X-rays, at various analyser crystal angles. Note the change in appearance of the contrast heterogeneities observed in the images throughout the analyser’s angular range. In the subsequent datasets, images were obtained by setting the analyser on the top of the rocking curve (unless specified otherwise) and at either 18 or 30 keV, as specified. Because it is not our intent to perform a complete analysis of the DEI application for cartilage nor is it possible to present all of the serial histological sections of each specimen, representative samples will be presented here. Examples of one normal talus and three tali with varying levels of grossly visible cartilage degeneration (all disarticulated from whole joints) are shown (bottom row) with their corresponding DE images (two top rows) in Fig. 4. The cartilage tissue can clearly be detected and distinguished from the bone. The DE image of the structure of cartilage on the normal talus looks homogeneous, with an average height of 1.5 mm and moderate density [Fig. 4(a),(e)]. This pattern changes in degenerated cartilage. The tissue is no longer homogeneous but shows patterns that suggest structural alterations [Fig. 4(b)–(d), (f)–(h)] that correspond to the degree and location of the grossly visualized degeneration. In particular, thin white lines can be seen running horizontally through the cartilage that correspond to the degree of roughening and fissuring of the cartilage surface in the degenerated specimens [arrows in Fig. 4(f)–(h)]. There is also a loss of contrast that increases as the degree of cartilage degeneration increases, especially at the most severely degenerated regions. The serial histological sections of these specimens substantiated the gross degeneration observed across the cartilage. The synchrotron X-ray and DE image of a portion of an osteoarthritic knee removed in the course of total knee replacement are shown in Fig. 5(a) and (b), respectively. At this degenerative stage, only small areas of the joint
166
J. Mollenhauer et al.: X-ray imaging of articular cartilage
Fig. 2. The ankle joint: an overview. (a) Medial aspect of the ankle joint from a left foot of a human skeleton. This joint is formed by the tibia and fibula articulating with the talus (arrow) to form the talocrural (ankle) joint. (b) The photograph shows the superior surface of the talus. The arrow indicates the orientation of the talus in relation to (c). (c) A DE image at 30 keV, with the X-ray beam parallel to the articular surface from posterior to anterior. The fine arrows indicate the bone/cartilage interface, with the cartilage (approx. 1.5 mm in height) as the less bright layer. The large arrow indicates the orientation of the DE image relative to the macroscopic image in (b). The actual resolution of the DE image is approximately 100 m and the image is approximately three-fold magnified as compared to the proportions of the talus. (d) Synchrotron radiograph of the same specimen as shown in (c).
Fig. 3. DEI images of articular cartilage along the rocking curve. An illustration showing the alterations in image appearance from articular cartilage as the analyser setting is taken through the rocking curve at 30 keV. The locations at which the images are taken are indicated on the rocking curve. Note the heterogeneities in contrast within the cartilage tissue whose appearance change at various points in the rocking curve.
surface are still covered with an intensely modified cartilage. Articular cartilage is barely visible in the synchrotron radiograph, while the DE image depicts a small 3 cm wide
and 2 cm segment of such residual cartilage on top of the femoral bone. Major horizontal contrast heterogeneities are visible in the DE image. A histological section of this
Osteoarthritis and Cartilage Vol. 10, No. 3
167
Fig. 4. Sections of DEI from the talar dome of the ankle joint. DEI of normal and damaged articular cartilage from the talus of a human ankle at 30 keV (top row) and 18 keV (middle row) with the corresponding macroscopic pictures (bottom row). The DEI cartilage tissue image of the intact talus (a),(e) shows a homogenous and moderately dense structure. In the damaged tali (b)–(d), (f)–(h) the degenerative sites (arrows) are clearly detected by distinctive heterogeneities within the images. In order to better visualize the damage in the photographs (i)–(l), the samples have been slightly rotated compared to the positioning for the DEI. The arrows indicate particular sites of lesions.
specimen [Fig. 5(c)] shows cartilage displaying severe signs of degeneration, including loss of proteoglycan staining and what appear to be structural changes. Serial sections from this specimen displayed similar structural and staining changes throughout. A representative view of a talar dome with its corresponding DE images, synchrotron radiograph and histology can be seen in Fig. 6. The articular cartilage was not visible without DEI as can be seen in the synchrotron radiograph in Fig. 6(d). However, the DE images at 18 keV [Fig. 6(b)] and 30 keV [Fig. 6(c)] clearly show the articular cartilage. The two different energy levels give rise to slightly different characteristics within the images, and the optimal energy level for articular cartilage has yet to be determined. However, it is apparent in DE images at both energies that the degenerated region of the talar cartilage gives rise to contrast heterogeneities and reduced contrast. A normallooking region of the specimen as depicted in the square on the left of the specimen in Fig. 6(c) can be seen in its Safranin O/fast green histological preparation in Fig. 6(e). Under polarized light, the same specimen displays a normal pattern of birefringence in the superficial and deep zones [Fig. 6(f)]. A region of the talar dome displaying a ‘bubbled’ region of cartilage (‘chondrophyte’) in which contrast heterogeneities (dark spots) can be seen is depicted in the square on the right side of the specimen of Fig. 6(c). Figure 6(g) is the Safranin O/fast green histological preparation of this chondrophyte region. The dark spot observed in the DE image in the right square appears, histologically, to be a vacuolated space (e.g. blister) in the cartilage surrounded by cartilage in a state of degeneration or remodeling. The polarized microscopic section of the same region [Fig. 6(h)] shows that the collagen fibers surrounding the ‘vacuolation’ have lost their normal birefringent pattern and thus display disorganization. Also, it can be noted that there is less DEI contrast in the region of the chondrophyte at the right in Fig. 6(b) and (c). Interestingly, this corresponds to a region of reduced Safranin O staining and loss of collagen birefringence patterns as depicted in
Fig. 6(g) and (h). Thus, it appears that the DE images reflect the normal-appearing and degenerated regions of the cartilage in these specimens. Although DEI is in its infancy in terms of development, an assessment of the potential of DEI for practical applicability in intact joints is warranted at this time. The DE images of the intact knee joint with all of its surrounding soft tissues, except skin, can be seen in Fig. 7. Figure 7(a) and (b) were taken with the analyser at the top of the rocking curve and at −3.6 rad, respectively. These are images of the medial femoral and tibial condyles with their associated articular hyaline cartilage and fibrocartilaginous menisci. The arrows point to the boundaries of the articular hyaline cartilage surfaces. Both the articular cartilage and menisci can be delineated even through the surrounding tissues. The white lines running diagonally through the cartilages appear to be the edges of the superimposed surrounding soft tissues of the joint as detected by the refraction capabilities of DEI.These two images, at 0 and −3.6 rad on the rocking curve, reflect the optimal scatter rejection and refraction qualities of the images, respectively. The articular cartilage was barely visible in the synchrotron radiograph of the same specimen. The articular cartilage of the intact ankle joint was visible in its DE image as well. The cartilage thickness could be measured, as the borders were apparent and the tissue appeared homogeneous (data not shown). Upon gross and histological examination, the cartilage was free of signs of degeneration.The articular cartilage was barely visible in the synchrotron radiograph of the same specimen.
Discussion The present study represents a first exploration into the possibilities of a novel X-ray technology, DEI, for cartilage imaging. Here, we have shown that visualization of articular cartilage of disarticulated and intact synovial joints is possible through DE imaging. It is not the intent of the present
168
Fig. 5. DEI and radiograph from a part of an osteoarthritic knee. A resected segment from a femoral condyle of a 67-year old patient who underwent total knee joint replacement due to end-stage OA is shown by synchrotron radiograph (a) and 18 keV-DEI (b) for comparison. In DEI, the residual cartilage is clearly visible and displays major contrast heterogeneities, as compared to the healthy cartilage of Fig. 2(c), and Fig. 4(a),(e). The histological section (c) demonstrates loss of Safranin O staining suggesting reduced proteoglycans within the depth of the cartilage.
study to compare the efficacy of DEI to other modes of cartilage imaging, rather, it is our purpose to introduce the imaging of cartilage, and some of its morphological pathologies, with X-rays for the first time. However, it may
J. Mollenhauer et al.: X-ray imaging of articular cartilage be mentioned here that because of the greater spatial resolution of X-ray radiography as compared to MRI at the present, theoretically, both different and greater detail should potentially be obtained through DEI technology. The application of DEI to mammography has already shown improved contrast of specific structures within human breast tissue in a clinical setting as compared to ordinary radiography21,22. In the present study we have shown that there is now good evidence for the potential of DEI to extend to skeletal radiology—not only for bone, but for cartilage as well. X-rays are ideal to evaluate changes in the subchondral bone, a major component of osteoarthritic disorders2,7,8. The additional information provided on articular cartilage through the DEI technique has the potential to revolutionize the utility of X-rays to monitor and diagnose OA. Importantly, at the same X-ray imaging energy, a comparable X-ray dose to that of conventional radiography is delivered to the specimen. In the present study, we have compared DE images of the sample specimens to their synchrotron radiographs. It should be noted that a synchrotron radiograph is a very high quality radiograph because of the monochromicity and collimation of the X-ray beam. Despite the high quality of ‘conventional’ synchrotron X-ray radiography, it is only capable of picking up absorption effects from a subject whereas DEI is capable of discerning the refraction and scatter rejection effects in addition to the absorption of conventional radiography. Ordinary radiography does not allow the visualization of articular cartilage let alone any defects within the cartilage. However, our results demonstrate that articular cartilage is visible on DE images of disarticulated and intact human ankle and knee joints. Furthermore, DEI has the potential to allow distinction between morphologically degenerated
Fig. 6. Cartilage of a talus demonstrating a degenerative site. (a) Photograph of the margin of a talar dome. The arrow points to a lesion or ‘chondrophyte’. (b) DE image of the lesion site at 18 keV. (c) DE image of the lesion site at 30 keV. The arrows in both (b) and (c) point to the cartilage/bone interface which appears to be interrupted at the far right in the 30 keV image in (c). (d) synchrotron radiograph of the specimen seen in (c). Note that the articular cartilage is not visible. (e) Safranin O stained section taken from a normal looking region seen in the square on the left in (c). (f) Picrosirius red stained histological section from the same region as viewed under polarizing light showing the normal birefringent pattern in the superficial and deep zones (*). (g) Safranin O staining of a histological section taken from the right square in (c). It is of interest that the lessened degree of contrast observed at the right corner of the image in (b) and (c) corresponds to regions of proteoglycan depletion as seen histologically. (h) Picrosirius red stained section of the chondrophyte region as viewed under polarizing light. Note the apparent space in the center and the lack of normal birefringence in the surrounding area indicating a disorientation of collagen fibers. Histological sections are magnified 4×.
Osteoarthritis and Cartilage Vol. 10, No. 3
169
Fig. 7. DE images of the medial condyle of an intact knee joint at the top of the rocking curve (a) and at −3.6 rad (b). The image was taken with all surrounding soft tissue, except the skin, in place. Note that the articular hyaline cartilage (whose borders can be seen at arrows) and menisci are visible even through the surrounding connective tissues.
and non-degenerated cartilages. Through DEI we have detected morphological defects in our study sample of disarticulated specimens at, and beneath, the cartilage surface that correspond to those of gross and histological examinations. Specifically, sites depicting states of degeneration such as fibrillation, more severe surface disruptions and even regions of cartilage loss beneath the surface were observable on DE images of the specimens. Of special interest are the thin white lines that appear in the DE images of the roughened, fissured regions of the specimens. It is possible that these white lines represent certain structural changes that give rise to large refraction detected by the imaging system. This would most probably develop at the edges of cartilage fibrillations, fissures or defects. However, condensed collagen fibrils may also cause such effects. The contrast may further be enhanced because the normally entrapped large proteoglycans are lost due to the damage of the collagen network23,24 and therefore the extinction of the X-ray beam is different in this area. It must be kept in mind that, because the DE images are a two-dimensional composite of characteristics throughout the tissue thickness being imaged, the images will display more structural information than provided by a single histological section. The DE images are, indeed, a reflection of the additive structural information provided by all of the serial sections from a specimen. Both the 18 and 30 keV energy levels used in the present study resulted in the visualization of articular cartilage, though they differ somewhat in appearance. 30 keV X-rays are of practical implication for patient imaging since they have sufficient penetrating power through intact human joints, while 18 keV energy is important for small animal subjects and serves as a basis of comparison since conventional radiographs taken at this energy have the optimum soft-tissue contrast. Further investigations into the energy level and positions on the rocking curve that provide the most representative images of the specimen as validated through gross, histological and biochemical analysis is underway. Variations in density, thickness and/or material properties refract the X-rays as they cross through the subject. These variations are generally in the submicroradian
range. The interaction of X-rays with a material object includes not only attenuation, but also refraction25–29 and ‘small angle scattering’. In particular, the small angle scattering effect (with an angular width of much less than one milliradian) carries information about an object’s structure on the length scale up to several m. Conventional radiography measures only the attenuation of X-rays in the object. Information about the X-ray refraction and small angle scattering is lost in conventional radiography due to the small angle nature of these effects. Likewise, microfocal radiography renders a high resolution, but only through absorption of X-rays using a micron-sized X-ray source. This technique, thus, provides fine detail of cancellous bone, but not of the cartilage30. However, DEI harnesses the information provided by small angle scattering (through the utilization of an analyser crystal) and is therefore capable of delineating adjacent tissues of similar densities, independent of tissue thickness. The character of this information varies depending upon the angular setting of the analyser crystal as can be seen from our cartilage images collected at various points on the so-called ‘rocking curve’. The optimal level of scatter rejection is attained at the peak of the rocking curve, whereas information on X-ray refraction is attained at the midpoints up and down the rocking curve. Thus, the superimposition of information from each point on the rocking curve can provide far more information than can be obtained with a single parameter (i.e. the absorption) as is the case with absorption radiography. This offers great potential for the assembly of multiple sets of data from a single specimen. At this stage, DEI utilizes a synchrotron source of Xrays. In principle, however, DEI is not dependent upon the synchrotron and further technology is under development to explore the applicability of DEI to ordinary X-ray systems. Although the use of DEI for the imaging of cartilage is in its very early developmental stages, the potential for this X-ray based system to image cartilage of not only disarticulated, but also intact joints is an obvious and important facility. For this reason, our ability to visualize the articular cartilage and fibrocartilaginous menisci of the intact knee joint with DEI in this study provides us with optimism for the future of DEI development for both research and clinical
170 settings. Furthermore, at different points in the rocking curve, structures can be delineated that have optimal scatter rejection or refraction qualities. Further away from the peak of the rocking curve, structures that result in the bending of X-rays will be particularly apparent as in case of the curvatures of the articular cartilage or menisci. Within the image of the intact knee joint, some structure of the superimposed joint-associated soft tissues are also visible. However, even through this superimposed tissue it should be possible to develop algorithms to measure the thickness of the articular cartilage and its surface intactness. With the eventual development of computed tomography for DEI (which is currently under way by Chapman and Zhong), there can be the partitioning of superimposed structures from one another. However, even in its current planar mode, DEI provides far more information concerning a subject’s absorption, refraction and scatter rejection properties than any other radiographic method. In summary, we have shown that with DEI technology articular cartilage of disarticulated and intact synovial joints can be radiographically imaged. Furthermore, gross cartilage defects and structural abnormalities, even at early stages of development, can be seen due to a combination of high spatial resolution and detection of X-ray refraction, extinction, and absorption patterns. We believe that DEI has the potential to be used as a component of a new X-ray generation in research and in clinical settings without requiring excessive exposure to X-rays31. This is especially applicable for skeletal radiology, and in areas where there is a need for simultaneous enhancement of soft tissue contrast. Although DEI currently utilizes synchrotron X-rays, in principle the technique is not dependent upon it. Future technological advances will carry the DEI methodology into a more practical setting. We are confident that this method may lead to an alternative approach to evaluate, diagnose and thereby influence the treatment of a disease that is one of the most important factors causing immobility within our society.
J. Mollenhauer et al.: X-ray imaging of articular cartilage
5. 6. 7. 8.
9. 10. 11. 12. 13.
14. 15.
16.
Acknowledgments The cartilage imaging carried out at the NSLS at Brookhaven National Laboratory by Matthias Aurich, M.D. was part of his post-doctorate training in the Department of Biochemistry at Rush Medical College and he contributed equally to the scientific content of the manuscript. We wish to thank Dr A. Valdellon and his staff at the Regional Organ Bank of Illinois for access to human donor cartilage. We also thank Dr Charles Peterfy for his helpful discussion and assistance with image analysis. This study was funded, in part, by NIH SCOR Grant #AR39239.
17. 18.
19.
20.
References 1. Wolbarst AB. In: Wolbast AB, Ed. Physics of Radiology. Connecticut: Appleton & Lange 1993: inside cover. 2. Kellgren JH, Lawrence JS. Radiologic assessment of osteoarthritis. Ann Rheum Dis 1957;16:494–502. 3. Rovati LC. Radiographic assessment. Introduction: existing methodology. Osteoarthritis Cart;1999: 427–9. 4. Buckland-Wright JC, Macfarlane DG, Lynch JA, Jasani MK, Bradshaw CR. Joint space width measures cartilage thickness in osteoarthritis of the knee: high resolution plain film and double contrast macro-
21.
22.
23.
radiographic investigation. Ann Rheum Dis 1995; 54:263–8. Hodler J, Resnik D. Current status of imaging of articular cartilage. Skeletal Radiol 1996;25:703–9. Hayes CW, Conway WF. Evaluation of articular cartilage: Radiographic and cross sectional imaging techniques. Radio Graphics 1992;12:409–28. Buckland-Wright C. Radiographic assessment of osteoarthritis: comparison between existing methodologies. Osteoarthritis Cart 1999;7:430–3. Vignon E, Conrozier T, Piperno M, Richard S, Carrillon Y, Fantino O. Radiographic assessment of hip and knee osteoarthritis. Recommendations: recommended guidelines. Osteoarthritis Cart 1999;7: 434–6. Adams ME, Wallace CJ. Quantitative imaging of osteoarthritis. Sem Arth Rheum 1991;20(6) (Suppl 2): 26–39. Blackburn WD, Chivers S, Bernreuter W. Cartilage imaging in osteoarthritis. Sem Arth Rheum 1996; 25(4):273–81. Waldschmidt JG, Braunstein EM, Buckwalter JA. Magnetic resonance imaging of osteoarthritis. Rheum Dis Clin North Am 1999;25(2):451–65. Disler DG, McCauley TR. Clinical magnetic resonance imaging of articular cartilage. Top Magn Res Imaging 1998;9(6): 360–76. Loeuille D, Olivier P, Mainard D, Gillet P, Netter P, Blum A. Review: Magnetic resonance imaging of normal and osteoarthritic cartilage. Arthritis Rheum 1998; 41(6):963–75. Chapman D, Thomlinson W, Johnston RE, Washburn D, Pisano E, Gmur N, et al. Diffraction enhanced X-ray imaging. Phys Med Biol 1997;42:2015–25. Zhong Z, Thomlinson W, Chapman D, Sayers D. Implementation of diffraction enhanced imaging experiments: at the NSLS and ASP. Nucl Instrum Meth Phys Res 2000;A447:556–68. Zachariasen WH. Theory of X-Ray Diffraction in Crystals. New York: John Wiley and Sons 1945. Guinier A, Fournet G, Walker CB, Yudowitch KL. Small Angle Scattering of X-rays. New York: John Wiley and Sons 1955. Muehleman C, Bareither D, Huch K, Cole A, Kuettner KE. Prevalence of degenerative morphological changes in the joints of the lower extremity. Osteoarthritis Cart 1997;5:1–15. Koepp H, Eger W, Muehleman C, Valdellon A, Buckwalter JA, Kuettner KE, et al. Prevalence of articular cartilage degeneration in the ankle and knee joints of human organ donors. J Orthop Sci 1999;4: 407–12. Rosenberg L. Chemical basis of the histological use of Safranin O in the study of articular cartilage. J Bone Joint Surg (AM) 1971;53:69–82. Chapman LD, Pisano E, Thomlinson W, Zhong Z, Johnston RE, Washburn D, et al. Medical applications of diffraction enhanced imaging. Breast Disease 1998;10:197–207. Pisano ED, Johnston RE, Chapman D, Geradts J, Iacocca MV, Livasy CA, et al. Human breast cancer specimens: Diffraction Enhanced Imaging with histologic correlation-improved conspicuity of lesion detail compared with digital radiography. Radiology 2000; 414:895–901. Perkins SJ, Nealis AS, Dunham DG, Hardingham TE, Muir IH. Molecular modeling of the multidomain
Osteoarthritis and Cartilage Vol. 10, No. 3
24. 25.
26. 27.
structures of the proteoglycan binding region and the link protein of cartilage by neutron and synchrotron X-ray scattering. Biochemistry 1991;30:10708–16. Kuettner KE, Aydelotte MB, Thonar E. Articular cartilage matrix and structure: a mini review. J Rheumatol 1991;18(Suppl 27):46–8. Ingal VN, Beliaevsky EA. X-ray plane-wave topography observation of phase contrast from a noncrystalline object. J Physics D Appl Phys 1995;28: 2314–17. Somenkov VA, Tkalich AK, Shil’shtein S. Refraction contrast in X-ray introscopy. Sov Phys Tech Phys 1991;36:1309–11. Podurets KM, Somenkov VA, Shil’shtein S. Refractioncontrast radiography. Sov Phys Tech Phys 1989;34: 654.
171 28. Gureyev TE, Wilkins SW. Regimes of X-ray phasecontrast imaging with perfect crystals. Il Nuovo Cimento 1997;19:545–52. 29. Bushuev VA, Ingal VN, Belyaevskaya EA. Dynamical theory of images generated by noncrystalline objects for the method of phase-dispersive introscopy. Crystallography Reports 1966;41:766–74. 30. Buckland-Wright JC, Lynch JA, Bird C. Microfocal techniques in quantitative radiography: measurement of cancellous bone organization. Br J Rheumatol 1996;35(Suppl 3):18–22. 31. Fitzgerald R. Phase-sensitive X-ray imaging: new approaches that can detect X-ray phase shifts within soft tissues show promise for clinical and biological applications. Physics Today 2000;53:23–36.