- Email: [email protected]
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND
MUSCULOSKELETAL ULTRASOUND
0033-8389/99 $8.00
+ .OO
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND Ronald S. Adler, PhD, MD
Significant advances in gray-scale and color flow ultrasound imaging have resulted in an expanded role of ultrasound in the evaluation of musculoskeletal path~logy.~, 9, l2 High-resolution imaging afforded by the current generation of high-frequency linear transducers has been shown to produce exquisite gray-scale images of tendons, muscles, nerves, and so forth?r9Likewise, improvements in color flow sensitivity allow the demonstration of alterations in blood flow associated with a variety of inflammatory, neoplastic, and posttraumatic states.I2 These improvements along with economic factors and availability have resulted in renewed interest in using ultrasound as a diagnostic tool. As such, the efficacy of ultrasound has been compared with other modalities, in particular MR imaging, as a cost-effective imaging alternative. This article does not illustrate the level of high-resolution ultrasound imaging attainable, but rather expands upon some new developments in ultrasound imaging, and also postulates new potential applications that further impact on the musculoskeletal system. This discussion falls into two general categories: (1) those developments that relate directly to imaging (i.e., three-dimensional imaging); and (2) those developments in which new information of a functional nature may be derived. In the latter case, quantitative analysis of
the gray-scale color flow or radio frequency data may provide new information unique to ultrasound. The importance of such developments is to help recognize the full potential of ultrasound imaging in the musculoskeletal system, and so to better define its role with respect to clinical diagnosis and management. NEW TISSUE APPLICATIONS
It is customary to think of ultrasound as a method only to visualize soft tissues. With respect to the musculoskeletal system, this entails characterization of tendons, muscles, bursae, synovial proliferation and so forth. The recognition that ultrasound can likewise characterize changes in the cortical surface of bone, or of the adjacent periosteum, has extended its applications to the diagnosis of fractures, osteomyelitis, and characterization of some neoplastic proce~ses.~ These have all been secondary applications in the sense that ultrasound is not the primary modality of choice to evaluate any of these processes. Alternatively, the presence of metallic hardware often precludes evaluation of the involved bones and adjacent soft tissues with CT or M R imaging. As such, ultrasound has shown promise in evaluation of prosthetic loosening or infection, in addition to being able to characterize other surrounding soft tissue struc-
From the Department of Radiology and Imaging, The Hospital for Special Surgery; and the Weil School of Medicine, Comell University, New York, New York
RADIOLOGIC CLINICS OF NORTH AMERICA VOLUME 37 NUMBER 4 JULY 1999
623
624
ADLER
tures. Of particular interest is the fact that metallic artifact has a unique sonographic signature due to a pronounced reverberation artifact, which readily allows it to be distinguished from other structures.22Likewise, high-density polyethylene is also characteristically different from other soft tissue structures and bone (Fig. 1). These features have suggested a role for ultrasound in the evaluation of polyethylene wear and prosthetic component loosening, resulting from wear-debris osteolysis or periprosthetic infection.
EXTENDED FIELD-OF-VIEW AND THREE-DIMENSIONAL TECHNIQUES
Tissue tracking and spatial registration have generated significant clinical and research interest in recent years by virtue of the appearance of extended field-of-view and three-dimensional imaging.*,4, 11, 19, 21 The former feature circumvents a major limitation of conventional gray-scale imaging, the small field-of-view. Whereas two-dimensional imaging suffices in most instances in making a
Figure 1. A, Lateral radiograph of a knee with a cemented total knee arthroplasty. The native tibia (t) and metallic tibia1 tray (m) are easily delineated. The polyethylene spacer (p) is seen as a poorly defined radiolucent zone situated immediately proximal to the metallic tray. 6, Sonogram obtained in parasagittal plane along anterior aspect of the same knee. The bone (t)/metal (m) interface is distinguished by the transition from weak to very strong reverberation artifact. The metal (m)/ polyethylene (p) transition appears as a sharp transition to a relatively echo-free band-like zone. The margin of this hypoechoic zone appears as a well-defined specular reflector caused by a relative large difference in its acoustic impedance with respect to soft tissue.
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND
diagnosis, extended field-of-view permits a better overall appreciation of a pathologic process relative to its surrounding tissues.2 The full extent of an abnormality can be placed on a single image, thereby allowing accurate assessment of size and effect on adjacent tissues (Fig. 2). Three-dimensional imaging permits multiplanar reformatting, enabling visualization of the optimal image plane, which may or may not be directly accessible. Imaging in the socalled C-plane, which may be thought of as a plane perpendicular to the insonating beam, provides an example of a nonaccessible image plane (Fig. 3). Alternatively, the capability simply to reformat the data removes some of the operator dependence inherent in ultrasound imaging by allowing one to review the image data from multiple effective scan angles. A variety of projection maps, (i.e., maximum pixel intensity) permit visualization of the entire three-dimensional data set, providing an additional method to better appreciate spatial relationships (Fig. 4). When combined with segmentation algorithms, which tend to be most amenable to color Doppler data, volumetric information may be obtained." Examples include total volume of a local muscle or tendon rupture, total vol-
625
ume of acute inflammatory pannus or total tumor vascularity estiaates, and so forth. All of these registration techniques require knowledge of the relative position of the transducer with respect to the image space in order to assign correctly a given two-dimensional image relative to its cohort. The composite data set can then be reprocessed to display a variety of features not encompassed by a single two-dimensional image. Electromagnetic and articulated arm methods that localize the transducer in space provide this capability but can be restrictive and, therefore, are of relatively limited application. Alternatively, techniques that utilize image speckle, which derive from subresolution scatterers and the finite system resolution, provide a more attractive method to achieve registration. Such techniques are generally referred to as speckle-tracking algorithms and rely on the extent to which images appear spatially correlated to place them appropriately in the image volume.4,l9 CORRELATION FEATURE IMAGING
Speckle-tracking offers much greater potential than simply a method to display a three-
Figure 2. Extended field-of-view, long axis view of supraspinatus tendon depicting the supraspinatus tendon from the acromion to a point beyond the tendon insertion onto the greater tuberosity. On a single image, an attenuated supraspinatus tendon (t) in which there is both bursa1 fraying and a distal full-thickness tear can be appreciated. Bursa1 thickening and a small acromial spur are present. In conventional imaging, the same information would require at least two images.
626
ADLER
Figure 3. A, Short axis view of a full-thickness supraspinatus tear. The lateral margins and undersurface of the tear are visible, but the full depth, cross-sectional area and quality of the retracted margins cannot be fully appreciated. B, C-plane images of the same supraspinatus tendon tear show the full area encompassed by the tear and the retracted tendon margins. A better appreciation of the tear's full extent is obtained. This image plane is equivalent to a bird's eye view of the rotator cuff.
dimensional volume or extend the field-of view. In order to register information spatially, we translate or rotate the transducer over some area of interest. Alternatively, observing a moving medium such as blood, while the transducer is in a fixed position provides an estimate of mean-tissue transit time. This method has been suggested as a means to derive true estimates of tissue perfusion, particularly when combined with ultrasound contrast agents.14Similarly, a tissue undergoing small elastic deformations can be compared with its undeformed state, resulting
in a differential tissue hardness, providing an alternative technique to evaluate low-contrast lesions or altered tissue states, such as relaxed and contracted muscle (Fig. 5).1° On gray-scale imaging muscles, tendons, and nerves display a property referred to as anisot~opy.~ This property relates to the varied appearance of the tissue as the angle of insonation is changed, and is a direct consequence of the orientation of extracellular components of the tissue (e.g., collagen in tendons). Loss of anisotropy in these tissues denotes underlying pathology. A scheme to quantify these
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND
627
terzone from alterations in acoustic attenuation has been shown to compare favorably with both micro CT and histology. CARTILAGE IMAGING
Figure 4. Maximum intensity projected image of vascularity in the biceps muscle following exercise. Good appreciation of vascular connectivity is obtained as well as the extent of vascularity.
It has been recognized for some time that ultrasound can detect gross changes in articular cartilage,' the limiting factor being acoustic access. In view of this limitation, global assessment of articular cartilage is best assessed with MR imaging. Alternatively, intraarticular ultrasound may provide the earliest imaging evidence of changes in the articular cartilage. Surface fibrillatory changes, an early manifestation of osteoarthritis, have been measured in vitro to accuracies of approximately 20 pm, taking advantage of the quantitative relationship of surface roughness and angle-dependent acoustic backscatter.5 Such techniques are expected to impact new developments in chondroprotective agents by permitting early detection of osteoarthritis and providing accurate assessment of disease progression. BLOOD FLOW IMAGING
changes accurately is desirable to determine progression or healing of a pre-existing abnormality. Two methods to quantify anisotropy have been s~ggested,'~, using statistically based correlation techniques. The more sensitive of these, in which local anisotropy index is calculated, is amenable to display as a parametric image (Fig. 6, and Color Plate, Fig. 1). BONE MINERALIZATION Broad-band acoustic attenuation and speedof-sound have been shown to display a quantitative relationship to minerali~ation.'~, 2o These methods correlate well with other existing techniques to determine bone mineral content. Qualitative assessment of bone formation across fractures and corticotomy defects has been demonstrated using gray-scale imaging, indicating the potential value of combining imaging with quantitative assessments of mineralizati~n.~~ Some progress in deriving an ultrasound-based quantitative measure of bone production has been demonstrated in dogs with Ilizarov distraction devices in s h 6 Calculation of the fibrous in-
A variety of symptomatic inflammatory and infectious lesions of the musculoskeletal system have been shown to manifest abnormally increased levels of vascularity on color flow imaging (Fig. 7 and Color Plate, Fig. 2): Power Doppler imaging has been suggested as a method to both detect and quantify these alterations in va~cu1arity.l~ This feature follows not only from improvements in sensitivity of power Doppler as compared with conventional color velocity maps, but also from the fact that power Doppler relates to the number of moving scatterers producing a Doppler shift in an insonated v01ume.~A consequence of this latter feature is that local tissue estimates of moving fractional blood volume may be obtained. Further, the combination of moving blood volume with estimates of rate of passage results in a local tissue perfusion measure. Such vascularity or perfusion measures provide a strong tool in assessing response to therapy. Preliminary data in patients with acute exacerbations of their inflammatory arthritides have shown a positive correlation of clinical improvement and reduced vascularity on power Doppler imaging.13 A correlation of
628
ADLER
Figure 5. Transversal matching in vivo sonograms/elastograms of a region of interest (40 x 50 mm2)from the quadriceps muscle group (rectus femoris and vastus group) at rest (upper panel) and during an isometric contraction of 10 Ib (lower panel). Imaging was done with a Diasonics Spectra clinical scanner that has been modified for elastographic imaging, using a 5-MHz, 40-mm linear array transducer. The subject was sitting on a comfortable examination bench with the left knee bent at a 90" angle, with the foot supported by a table adjacent to a 15-lb force gauge. The foot was allowed to voluntarily slide (with reduced friction) on a supporting table and pressed against the force gauge. The subject was asked to slide his left foot against the force gauge to generate the desired force of isometric contraction. Observe that elastography clearly demonstrates the change in muscle contraction level. In this patient, the elastogram of the lower panel clearly demonstrates the increased contraction level of the rectus femoris muscle. RF = rectus femoris; VL = vastus lateralis; VM = vastus medialis; F = femur. (From Kallel F, et al: Elastographic imaging of human muscle during isometric contraction. Ultrasound Med Biol (in press); with permission.)
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND
629
Figure 6. The anisotrophic echotexture of tendon is quantified in the spatial frequency domain. Using a moving 32 x 32 pixel window, the two-dimensional spectrum is computed, and the Eigen values of the covariance matrix are used to denote the axes of the local ellipsoidal spectra. The major-to-minor axis ratio (AR) can then be used to separate out normal tendon. A directionality vector also can be assigned to the angle of the major axis. The top figure shows a normal Achilles’ tendon, and the middle figure shows the corresponding AR segmented image. The directionality image (boffom)shows the strong coherence pattern in the tendon. Ultrasound Med Biol; (From Tuthill TA: Frequency analysis of echotexture in tendon. Ultrasound Med Biol (in press); with permission.) See Color Plate 1, Fig. 1.
increased power Doppler signal and increased vascularity has been demonstrated histologically in an animal model of septic arthritis.16 Further, estimates of fractional moving blood volume have been demonstrated to correlate well with local tissue perfusion in an animal model, using colorlabeled microspheres as the gold standard (Forsberg F, private communication). The fact that such a quantitative relationship should exist appears to be well-grounded biochemically. In recent years, it has became apparent that a variety of so-called angiogenic-sensitive neoplasms (e.g., invasive breast carcinoma) bear a quantitative relationship between microvessel density and biologic activity. Likewise, the release of angiogenic factors has
been implicated in inflammatory arthritides, such as rheumatoid and psoriatic arthritis.8 SUMMARY
Improvements in high-resolution gray-scale imaging and clinical expertise performing musculoskeletal ultrasound will undoubtedly continue. Development of digital beam formers, two-dimensional arrays along with exploitation of nonlinear techniques to achieve higher resolution and use of ultrasound contrast to improve flow sensitivity will all contribute to the utility of ultrasound in the musculoskeletal system. It behooves the radiologic community to become familiar with
630
ADLER
Figure 7. Power Doppler image of quadriceps muscle group in patient with pyomyositis. A discrete abscess is present (arrow) surrounded by surrounding rim of increased blood flow. The appearance is analogous to rim enhancement seen in either CT or MR imaging following contrast administration. See Color Plate 1, Fig. 2.
these techniques, not only for economic reasons, but also because of the rich complement of future applications of this modality. The few potential applications mentioned here may only scratch the surface of what is possible. In addition to improved images of tissue morphology, ultrasound may play a role in functional and quantitative assessment of soft tissues. It may likewise play a role in the evaluation of prosthetic implants, bone mineralization, and cartilage integrity. Thus, the role of this modality in future musculoskeletal applications may significantly impact clinical diagnosis and therapy. References 1. Aisen A, McCune WA, Martel WM: Sonographic evaluation of the cartilage of the knee. Radiology 153781-784,1984 2. Barberie JE, Wong AD, Cooperberb PL, Carson BW Extended field-of-view sonography in musculoskeletal disorders. AJR Am J Roentgenol 171:751-757,1998 3. Bouffard JA, Eyler WR, Introcaso JH, van Holsbeeck M: Sonography of tendons. Ultrasound Quarterly 11~259-286,1993 4. Chen J-F, Fowlkes JB, Carson PL, Rubin J M Determination of scan-plane motion using speckle decorrelation: Theoretical considerations and initial test. Int J Imaging Syst Techno1 8:38-44,1996
5. Chiang EH, Laing TJ, Meyers CR, et a1 Ultrasonic characterization of in vitro osteoarthritic articular cartilage with validation by confocal microscopy. U1trasound Med Biol23205213,1997 6. Daniel BL, Nicholas A, Waanders MS, et al: The use of ultrasound mean acoustic attenuation to quantify bone formation during distraction osteogenesis performed by the Ilizarov method. Invest Radio1 29:933939, 1994 7. Dymling SO, Persson HW, Hertz C H Measurement of blood perfusion in tissue using Doppler sonography. Ultrasound Med Biol 17433-444, 1991 8. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other diseases. Nature Med 1:27-31, 1995 9. Fornage BD Musculoskeletal Ultrasound. New York, Churchill Livingstone, 1995 10. Kallel F, Tan FC, Krouskop T, Ophir J: Elastrographic imaging of human muscle during isometric contraction Ultrasound Med Biol (in press) 11. Moskalik A, Carson PL, Meyer CR, et al: Registration of three dimensional ultrasound scans of the breast for refraction and motion correction. Ultrasound Med Biol 21:769-778, 1995 12. Newman JS, Adler RS, Bude RO, Rubin JM: Detection of soft tissue hyperemia: Value of power Doppler sonography. AJR Am J Roentgenol 163385389,1994 13. Newman JS, Laing TJ, McCarthy CJ, Adler Rs: Power Doppler sonography in synovitis: Assessment of therapeutic response-preliminary observations. Radi010a 198582-584, 1996 14. Rubin JM, Adler RS, Fowlkes JB, et al: Fractional moving blood volume estimation using power Doppler imaging. Radiology 197183-190, 1995 15. Rubin JM, Fowlkes JB, Bude RO, et al: Correlation
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND length A potential new technique for evaluating tendon abnormalities. J Ultrasound Med 16S90, 1998 16. Strouse PJ, DiPietro MA, Teo EHJ, et al: Power Doppler evaluation of joint effusions: Investigations in a rabbit model. J Pediatr Radio1 (in press) 17. Tavacoli MB, Evans J A Dependence of velocity and attenuation of ultrasound in bone on the mineral content. Phys Med Biol361529-1537,1991 18. Tuthill T, Fowlkes JB,Rubin JM, Carson PL Automatic tendon segmentation based o n spacial frequency movements. J Ultrasound Med 16S16,1998 19. Wagner RF, Insana MF, Brown DG: Statistical properties of radio-frequency and envelope-detected signals with application to medical ultrasound. J Opt Soc Am 4910-922,1987
631
20. Waud CE, Lew R, Baran DT: The relationship between ultrasound and densitometric measurements of bone mass at the calcaneus of women. Calcif Tissue Int 51:415-418,1992 21. Weng L, Tirumalai AP, Lowery CM, et al: Extended field-of-view imaging technology. Radiology 203877880, 1997 22. Yasher AA, Adler RS, Grady-Benson JC, et a1 Ultrasound method to evaluate polyethylene component wear in total knee replacement arthroplasty. Am J Orthop 25:702-704,1996 23. Young JWR, Kostrubiak IS, Resnik C, et al: Sonographic evaluation of bone production at the distraction site of Ilizrarov limb-lengthening procedures. AJR Am J Roentgen01 154:125-128, 1990
Address reprint requests to Ronald S . Adler, PhD, MD The Hospital for Special Surgery 535 East 70th Street New York, NY 10021 e-mail: [email protected]
0033-8389/99 $8.00
+ .OO
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND Ronald S. Adler, PhD, MD
Significant advances in gray-scale and color flow ultrasound imaging have resulted in an expanded role of ultrasound in the evaluation of musculoskeletal path~logy.~, 9, l2 High-resolution imaging afforded by the current generation of high-frequency linear transducers has been shown to produce exquisite gray-scale images of tendons, muscles, nerves, and so forth?r9Likewise, improvements in color flow sensitivity allow the demonstration of alterations in blood flow associated with a variety of inflammatory, neoplastic, and posttraumatic states.I2 These improvements along with economic factors and availability have resulted in renewed interest in using ultrasound as a diagnostic tool. As such, the efficacy of ultrasound has been compared with other modalities, in particular MR imaging, as a cost-effective imaging alternative. This article does not illustrate the level of high-resolution ultrasound imaging attainable, but rather expands upon some new developments in ultrasound imaging, and also postulates new potential applications that further impact on the musculoskeletal system. This discussion falls into two general categories: (1) those developments that relate directly to imaging (i.e., three-dimensional imaging); and (2) those developments in which new information of a functional nature may be derived. In the latter case, quantitative analysis of
the gray-scale color flow or radio frequency data may provide new information unique to ultrasound. The importance of such developments is to help recognize the full potential of ultrasound imaging in the musculoskeletal system, and so to better define its role with respect to clinical diagnosis and management. NEW TISSUE APPLICATIONS
It is customary to think of ultrasound as a method only to visualize soft tissues. With respect to the musculoskeletal system, this entails characterization of tendons, muscles, bursae, synovial proliferation and so forth. The recognition that ultrasound can likewise characterize changes in the cortical surface of bone, or of the adjacent periosteum, has extended its applications to the diagnosis of fractures, osteomyelitis, and characterization of some neoplastic proce~ses.~ These have all been secondary applications in the sense that ultrasound is not the primary modality of choice to evaluate any of these processes. Alternatively, the presence of metallic hardware often precludes evaluation of the involved bones and adjacent soft tissues with CT or M R imaging. As such, ultrasound has shown promise in evaluation of prosthetic loosening or infection, in addition to being able to characterize other surrounding soft tissue struc-
From the Department of Radiology and Imaging, The Hospital for Special Surgery; and the Weil School of Medicine, Comell University, New York, New York
RADIOLOGIC CLINICS OF NORTH AMERICA VOLUME 37 NUMBER 4 JULY 1999
623
624
ADLER
tures. Of particular interest is the fact that metallic artifact has a unique sonographic signature due to a pronounced reverberation artifact, which readily allows it to be distinguished from other structures.22Likewise, high-density polyethylene is also characteristically different from other soft tissue structures and bone (Fig. 1). These features have suggested a role for ultrasound in the evaluation of polyethylene wear and prosthetic component loosening, resulting from wear-debris osteolysis or periprosthetic infection.
EXTENDED FIELD-OF-VIEW AND THREE-DIMENSIONAL TECHNIQUES
Tissue tracking and spatial registration have generated significant clinical and research interest in recent years by virtue of the appearance of extended field-of-view and three-dimensional imaging.*,4, 11, 19, 21 The former feature circumvents a major limitation of conventional gray-scale imaging, the small field-of-view. Whereas two-dimensional imaging suffices in most instances in making a
Figure 1. A, Lateral radiograph of a knee with a cemented total knee arthroplasty. The native tibia (t) and metallic tibia1 tray (m) are easily delineated. The polyethylene spacer (p) is seen as a poorly defined radiolucent zone situated immediately proximal to the metallic tray. 6, Sonogram obtained in parasagittal plane along anterior aspect of the same knee. The bone (t)/metal (m) interface is distinguished by the transition from weak to very strong reverberation artifact. The metal (m)/ polyethylene (p) transition appears as a sharp transition to a relatively echo-free band-like zone. The margin of this hypoechoic zone appears as a well-defined specular reflector caused by a relative large difference in its acoustic impedance with respect to soft tissue.
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND
diagnosis, extended field-of-view permits a better overall appreciation of a pathologic process relative to its surrounding tissues.2 The full extent of an abnormality can be placed on a single image, thereby allowing accurate assessment of size and effect on adjacent tissues (Fig. 2). Three-dimensional imaging permits multiplanar reformatting, enabling visualization of the optimal image plane, which may or may not be directly accessible. Imaging in the socalled C-plane, which may be thought of as a plane perpendicular to the insonating beam, provides an example of a nonaccessible image plane (Fig. 3). Alternatively, the capability simply to reformat the data removes some of the operator dependence inherent in ultrasound imaging by allowing one to review the image data from multiple effective scan angles. A variety of projection maps, (i.e., maximum pixel intensity) permit visualization of the entire three-dimensional data set, providing an additional method to better appreciate spatial relationships (Fig. 4). When combined with segmentation algorithms, which tend to be most amenable to color Doppler data, volumetric information may be obtained." Examples include total volume of a local muscle or tendon rupture, total vol-
625
ume of acute inflammatory pannus or total tumor vascularity estiaates, and so forth. All of these registration techniques require knowledge of the relative position of the transducer with respect to the image space in order to assign correctly a given two-dimensional image relative to its cohort. The composite data set can then be reprocessed to display a variety of features not encompassed by a single two-dimensional image. Electromagnetic and articulated arm methods that localize the transducer in space provide this capability but can be restrictive and, therefore, are of relatively limited application. Alternatively, techniques that utilize image speckle, which derive from subresolution scatterers and the finite system resolution, provide a more attractive method to achieve registration. Such techniques are generally referred to as speckle-tracking algorithms and rely on the extent to which images appear spatially correlated to place them appropriately in the image volume.4,l9 CORRELATION FEATURE IMAGING
Speckle-tracking offers much greater potential than simply a method to display a three-
Figure 2. Extended field-of-view, long axis view of supraspinatus tendon depicting the supraspinatus tendon from the acromion to a point beyond the tendon insertion onto the greater tuberosity. On a single image, an attenuated supraspinatus tendon (t) in which there is both bursa1 fraying and a distal full-thickness tear can be appreciated. Bursa1 thickening and a small acromial spur are present. In conventional imaging, the same information would require at least two images.
626
ADLER
Figure 3. A, Short axis view of a full-thickness supraspinatus tear. The lateral margins and undersurface of the tear are visible, but the full depth, cross-sectional area and quality of the retracted margins cannot be fully appreciated. B, C-plane images of the same supraspinatus tendon tear show the full area encompassed by the tear and the retracted tendon margins. A better appreciation of the tear's full extent is obtained. This image plane is equivalent to a bird's eye view of the rotator cuff.
dimensional volume or extend the field-of view. In order to register information spatially, we translate or rotate the transducer over some area of interest. Alternatively, observing a moving medium such as blood, while the transducer is in a fixed position provides an estimate of mean-tissue transit time. This method has been suggested as a means to derive true estimates of tissue perfusion, particularly when combined with ultrasound contrast agents.14Similarly, a tissue undergoing small elastic deformations can be compared with its undeformed state, resulting
in a differential tissue hardness, providing an alternative technique to evaluate low-contrast lesions or altered tissue states, such as relaxed and contracted muscle (Fig. 5).1° On gray-scale imaging muscles, tendons, and nerves display a property referred to as anisot~opy.~ This property relates to the varied appearance of the tissue as the angle of insonation is changed, and is a direct consequence of the orientation of extracellular components of the tissue (e.g., collagen in tendons). Loss of anisotropy in these tissues denotes underlying pathology. A scheme to quantify these
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND
627
terzone from alterations in acoustic attenuation has been shown to compare favorably with both micro CT and histology. CARTILAGE IMAGING
Figure 4. Maximum intensity projected image of vascularity in the biceps muscle following exercise. Good appreciation of vascular connectivity is obtained as well as the extent of vascularity.
It has been recognized for some time that ultrasound can detect gross changes in articular cartilage,' the limiting factor being acoustic access. In view of this limitation, global assessment of articular cartilage is best assessed with MR imaging. Alternatively, intraarticular ultrasound may provide the earliest imaging evidence of changes in the articular cartilage. Surface fibrillatory changes, an early manifestation of osteoarthritis, have been measured in vitro to accuracies of approximately 20 pm, taking advantage of the quantitative relationship of surface roughness and angle-dependent acoustic backscatter.5 Such techniques are expected to impact new developments in chondroprotective agents by permitting early detection of osteoarthritis and providing accurate assessment of disease progression. BLOOD FLOW IMAGING
changes accurately is desirable to determine progression or healing of a pre-existing abnormality. Two methods to quantify anisotropy have been s~ggested,'~, using statistically based correlation techniques. The more sensitive of these, in which local anisotropy index is calculated, is amenable to display as a parametric image (Fig. 6, and Color Plate, Fig. 1). BONE MINERALIZATION Broad-band acoustic attenuation and speedof-sound have been shown to display a quantitative relationship to minerali~ation.'~, 2o These methods correlate well with other existing techniques to determine bone mineral content. Qualitative assessment of bone formation across fractures and corticotomy defects has been demonstrated using gray-scale imaging, indicating the potential value of combining imaging with quantitative assessments of mineralizati~n.~~ Some progress in deriving an ultrasound-based quantitative measure of bone production has been demonstrated in dogs with Ilizarov distraction devices in s h 6 Calculation of the fibrous in-
A variety of symptomatic inflammatory and infectious lesions of the musculoskeletal system have been shown to manifest abnormally increased levels of vascularity on color flow imaging (Fig. 7 and Color Plate, Fig. 2): Power Doppler imaging has been suggested as a method to both detect and quantify these alterations in va~cu1arity.l~ This feature follows not only from improvements in sensitivity of power Doppler as compared with conventional color velocity maps, but also from the fact that power Doppler relates to the number of moving scatterers producing a Doppler shift in an insonated v01ume.~A consequence of this latter feature is that local tissue estimates of moving fractional blood volume may be obtained. Further, the combination of moving blood volume with estimates of rate of passage results in a local tissue perfusion measure. Such vascularity or perfusion measures provide a strong tool in assessing response to therapy. Preliminary data in patients with acute exacerbations of their inflammatory arthritides have shown a positive correlation of clinical improvement and reduced vascularity on power Doppler imaging.13 A correlation of
628
ADLER
Figure 5. Transversal matching in vivo sonograms/elastograms of a region of interest (40 x 50 mm2)from the quadriceps muscle group (rectus femoris and vastus group) at rest (upper panel) and during an isometric contraction of 10 Ib (lower panel). Imaging was done with a Diasonics Spectra clinical scanner that has been modified for elastographic imaging, using a 5-MHz, 40-mm linear array transducer. The subject was sitting on a comfortable examination bench with the left knee bent at a 90" angle, with the foot supported by a table adjacent to a 15-lb force gauge. The foot was allowed to voluntarily slide (with reduced friction) on a supporting table and pressed against the force gauge. The subject was asked to slide his left foot against the force gauge to generate the desired force of isometric contraction. Observe that elastography clearly demonstrates the change in muscle contraction level. In this patient, the elastogram of the lower panel clearly demonstrates the increased contraction level of the rectus femoris muscle. RF = rectus femoris; VL = vastus lateralis; VM = vastus medialis; F = femur. (From Kallel F, et al: Elastographic imaging of human muscle during isometric contraction. Ultrasound Med Biol (in press); with permission.)
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND
629
Figure 6. The anisotrophic echotexture of tendon is quantified in the spatial frequency domain. Using a moving 32 x 32 pixel window, the two-dimensional spectrum is computed, and the Eigen values of the covariance matrix are used to denote the axes of the local ellipsoidal spectra. The major-to-minor axis ratio (AR) can then be used to separate out normal tendon. A directionality vector also can be assigned to the angle of the major axis. The top figure shows a normal Achilles’ tendon, and the middle figure shows the corresponding AR segmented image. The directionality image (boffom)shows the strong coherence pattern in the tendon. Ultrasound Med Biol; (From Tuthill TA: Frequency analysis of echotexture in tendon. Ultrasound Med Biol (in press); with permission.) See Color Plate 1, Fig. 1.
increased power Doppler signal and increased vascularity has been demonstrated histologically in an animal model of septic arthritis.16 Further, estimates of fractional moving blood volume have been demonstrated to correlate well with local tissue perfusion in an animal model, using colorlabeled microspheres as the gold standard (Forsberg F, private communication). The fact that such a quantitative relationship should exist appears to be well-grounded biochemically. In recent years, it has became apparent that a variety of so-called angiogenic-sensitive neoplasms (e.g., invasive breast carcinoma) bear a quantitative relationship between microvessel density and biologic activity. Likewise, the release of angiogenic factors has
been implicated in inflammatory arthritides, such as rheumatoid and psoriatic arthritis.8 SUMMARY
Improvements in high-resolution gray-scale imaging and clinical expertise performing musculoskeletal ultrasound will undoubtedly continue. Development of digital beam formers, two-dimensional arrays along with exploitation of nonlinear techniques to achieve higher resolution and use of ultrasound contrast to improve flow sensitivity will all contribute to the utility of ultrasound in the musculoskeletal system. It behooves the radiologic community to become familiar with
630
ADLER
Figure 7. Power Doppler image of quadriceps muscle group in patient with pyomyositis. A discrete abscess is present (arrow) surrounded by surrounding rim of increased blood flow. The appearance is analogous to rim enhancement seen in either CT or MR imaging following contrast administration. See Color Plate 1, Fig. 2.
these techniques, not only for economic reasons, but also because of the rich complement of future applications of this modality. The few potential applications mentioned here may only scratch the surface of what is possible. In addition to improved images of tissue morphology, ultrasound may play a role in functional and quantitative assessment of soft tissues. It may likewise play a role in the evaluation of prosthetic implants, bone mineralization, and cartilage integrity. Thus, the role of this modality in future musculoskeletal applications may significantly impact clinical diagnosis and therapy. References 1. Aisen A, McCune WA, Martel WM: Sonographic evaluation of the cartilage of the knee. Radiology 153781-784,1984 2. Barberie JE, Wong AD, Cooperberb PL, Carson BW Extended field-of-view sonography in musculoskeletal disorders. AJR Am J Roentgenol 171:751-757,1998 3. Bouffard JA, Eyler WR, Introcaso JH, van Holsbeeck M: Sonography of tendons. Ultrasound Quarterly 11~259-286,1993 4. Chen J-F, Fowlkes JB, Carson PL, Rubin J M Determination of scan-plane motion using speckle decorrelation: Theoretical considerations and initial test. Int J Imaging Syst Techno1 8:38-44,1996
5. Chiang EH, Laing TJ, Meyers CR, et a1 Ultrasonic characterization of in vitro osteoarthritic articular cartilage with validation by confocal microscopy. U1trasound Med Biol23205213,1997 6. Daniel BL, Nicholas A, Waanders MS, et al: The use of ultrasound mean acoustic attenuation to quantify bone formation during distraction osteogenesis performed by the Ilizarov method. Invest Radio1 29:933939, 1994 7. Dymling SO, Persson HW, Hertz C H Measurement of blood perfusion in tissue using Doppler sonography. Ultrasound Med Biol 17433-444, 1991 8. Folkman J: Angiogenesis in cancer, vascular, rheumatoid and other diseases. Nature Med 1:27-31, 1995 9. Fornage BD Musculoskeletal Ultrasound. New York, Churchill Livingstone, 1995 10. Kallel F, Tan FC, Krouskop T, Ophir J: Elastrographic imaging of human muscle during isometric contraction Ultrasound Med Biol (in press) 11. Moskalik A, Carson PL, Meyer CR, et al: Registration of three dimensional ultrasound scans of the breast for refraction and motion correction. Ultrasound Med Biol 21:769-778, 1995 12. Newman JS, Adler RS, Bude RO, Rubin JM: Detection of soft tissue hyperemia: Value of power Doppler sonography. AJR Am J Roentgenol 163385389,1994 13. Newman JS, Laing TJ, McCarthy CJ, Adler Rs: Power Doppler sonography in synovitis: Assessment of therapeutic response-preliminary observations. Radi010a 198582-584, 1996 14. Rubin JM, Adler RS, Fowlkes JB, et al: Fractional moving blood volume estimation using power Doppler imaging. Radiology 197183-190, 1995 15. Rubin JM, Fowlkes JB, Bude RO, et al: Correlation
FUTURE AND NEW DEVELOPMENTS IN MUSCULOSKELETAL ULTRASOUND length A potential new technique for evaluating tendon abnormalities. J Ultrasound Med 16S90, 1998 16. Strouse PJ, DiPietro MA, Teo EHJ, et al: Power Doppler evaluation of joint effusions: Investigations in a rabbit model. J Pediatr Radio1 (in press) 17. Tavacoli MB, Evans J A Dependence of velocity and attenuation of ultrasound in bone on the mineral content. Phys Med Biol361529-1537,1991 18. Tuthill T, Fowlkes JB,Rubin JM, Carson PL Automatic tendon segmentation based o n spacial frequency movements. J Ultrasound Med 16S16,1998 19. Wagner RF, Insana MF, Brown DG: Statistical properties of radio-frequency and envelope-detected signals with application to medical ultrasound. J Opt Soc Am 4910-922,1987
631
20. Waud CE, Lew R, Baran DT: The relationship between ultrasound and densitometric measurements of bone mass at the calcaneus of women. Calcif Tissue Int 51:415-418,1992 21. Weng L, Tirumalai AP, Lowery CM, et al: Extended field-of-view imaging technology. Radiology 203877880, 1997 22. Yasher AA, Adler RS, Grady-Benson JC, et a1 Ultrasound method to evaluate polyethylene component wear in total knee replacement arthroplasty. Am J Orthop 25:702-704,1996 23. Young JWR, Kostrubiak IS, Resnik C, et al: Sonographic evaluation of bone production at the distraction site of Ilizrarov limb-lengthening procedures. AJR Am J Roentgen01 154:125-128, 1990
Address reprint requests to Ronald S . Adler, PhD, MD The Hospital for Special Surgery 535 East 70th Street New York, NY 10021 e-mail: [email protected]