| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology and Radiological Science, Johns Hopkins Medical
Institutions, 601 N Caroline St., JHOC 3171C, Baltimore, MD 21287.
2 Department of Orthopaedic Surgery, Johns Hopkins Medical Institutions,
Baltimore, MD 21287.
Received June 16, 2004;
accepted after revision November 10, 2004.
Address correspondence to L. M. Fayad
(lfayad1{at}jhmi.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
MATERIALS AND METHODS. Two reviewers retrospectively reviewed 45 MR images, 37 CT scans, and 43 radiographs in 59 patients (30 biopsy-proven pathologic fractures and 29 stress fractures followed to resolution). The features observed on MRI were abnormal bone marrow (well-defined, ill-defined); intracortical, periosteal, or muscle T1 or T2 signal; endosteal scalloping; and a soft-tissue mass. The features seen on CT were marrow abnormality and character (well-defined, ill-defined, permeative, moth-eaten), endosteal scalloping, periosteal reaction (benign, aggressive), and a soft-tissue mass. Reviewers rated their confidence for diagnosing a pathologic fracture on a 1-3 scale (< 50%, 50-95%, > 95% sure, respectively) with each technique. Performance of each technique was defined by reviewer accuracy and area under the receiver operating characteristic curve (Az); the frequency with which the MRI and CT features were associated with pathologic and stress fractures was calculated.
RESULTS. For both reviewers, accuracy for differentiating pathologic from stress fractures was highest on MRI (accuracy/Az: reviewer 1, 98%/0.97; reviewer 2, 93%/0.99); CT (reviewer 1, 88%/0.83; reviewer 2, 82%/0.90) was less accurate than radiography (reviewer 1, 94%/0.98; reviewer 2, 88%/0.96). On MRI, pathologic fractures compared with stress fractures exhibited well-defined T1 marrow signal (83% vs 7%, respectively; p < 0.001), endosteal scalloping (58% vs 0%, p < 0.001), muscle signal (83% vs 48%, p = 0.026), and a soft-tissue mass (67% vs 0%, p < 0.001). On CT, pathologic fractures compared with stress fractures exhibited marrow abnormality (84% vs 17%, respectively; p = 0.001), endosteal scalloping (44% vs 0%, p = 0.006), and aggressive periosteal reaction (36% vs 0%, p = 0.04).
CONCLUSION. MRI is useful for distinguishing pathologic from stress fractures, especially after inconclusive radiographic findings. Specifically, pathologic fractures exhibit well-defined T1 marrow alterations, endosteal scalloping, and adjacent soft-tissue abnormalities.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Although there are case series describing the appearance of stress fractures with cross-sectional CT and MRI [7, 11-21], to our knowledge, characteristics differentiating these fractures from pathologic fractures have been addressed in a very limited fashion with regard to long bones [9]. However, there has been recent emphasis on distinguishing vertebral pathologic fractures from stress fractures using advanced MRI techniques [22, 23].
The purposes of this report are to describe the confidence with which radiologists discriminate stress fractures from pathologic fractures in long bones using radiography, CT, and conventional MRI and to further define cross-sectional imaging features that distinguish these entities.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Hypotheses of the Study
The hypotheses of this study were, first, that cross-sectional imaging aids
the radiologist in detecting pathologic fractures with greater certainty than
radiography; second, that features exist on CT and MRI that accurately
distinguish pathologic fractures from stress fractures of long bones; and,
third, that MRI is superior to CT for the identification of pathologic
fractures given its increased sensitivity to bone marrow abnormalities
[24].
Acquisition of Studies and Patient Population
Using a computerized imaging database, we retrospectively identified 59
patients who had undergone CT or MRI with reported fractures of long bones
between September 1997 and December 2003. Subsequently, a review of these
patients' medical charts for history of biopsy and clinical follow-up was
performed. Only biopsy-proven cases or those with typical clinical outcome
were included. For pathologic fractures, 28 cases were biopsy-proven. In the
remaining two cases, one patient had widespread myeloma diagnosed by bone
marrow aspirate and one patient had prostate carcinoma, multiple bone lesions,
and prostate-specific antigen level of 624.7 ng/mL. For stress fractures,
eight cases were biopsy-proven to be negative for tumor. The remaining cases
consisted of patients who were followed radiographically or clinically to
resolution of the fracture.
The final patient population was composed of 59 patients: 30 with pathologic fractures and 29 with stress fractures. In this patient population, a total of 31 radiographs, 25 CT scans, and 24 MR images were obtained in the 30 patients with pathologic fractures; 12 radiographs, 12 CT scans, and 21 MR images were obtained in the patients with stress fractures. Two patients had bilateral pathologic femur fractures; hence, the total number of pathologic fractures was 32 in 30 patients.
Radiography was performed the same day as or within 24 days of cross-sectional imaging. Twelve patients had both CT and MRI examinations: In three, imaging was performed the same day; in three, within 1 day; in three, within 5 days; in two, within 14 days; and in one, within 24 days.
The patient group was composed of 29 females (11 with pathologic fractures and 18 with stress fractures) and 30 males (18 with pathologic fractures and 12 with stress fractures). The mean age of the patients with pathologic fractures was 47.7 years (range, 6-96 years) and that for the patients with fatigue or insufficiency fractures was 52 years (range, 2-91 years). Four patients with stress fractures had a history of malignancy.
Pathologic fractures included six benign lesions: two nonossifying fibromas, one fibrous dysplasia, one enchondroma, one eosinophilic granuloma, one unicameral bone cyst. The remaining 24 malignant lesions included seven sarcomas (one osteogenic sarcoma, one leiomyosarcoma, three malignant fibrous histiocytomas, one Ewing's sarcoma, one chondrosarcoma), five lymphomas, 10 adenocarcinomas (four breast, three lung, one esophageal, one prostate, one thyroid), and two cases of myeloma.
Radiography Examinations
Imaging was performed in standard radiographic projections, usually
consisting of two views of the bone in question.
CT Examinations
CT was performed using either a Somatom Plus 4 scanner (Siemens Medical
Solutions) or a VolumeZoom unit (Siemens Medical Solutions). An unenhanced
helical study of the limb in question was performed using 1- to 3-mm slice
thickness, 1-2 pitch, 1- to 3-mm reconstruction interval, 280 mA, and 120 kVp.
Subsequent sagittal and coronal reconstructed images were obtained. In 10
cases, an examination of the limbs was included as part of a dedicated CT
examination of the chest or abdomen using 5-mm slice thickness, pitch of 4,
5-mm reconstruction interval, 280 mA, and 120 kVp. In these cases, neither
sagittal nor coronal reconstructed imaging was performed.
In seven cases, a contrast-enhanced examination was then performed after the IV power injection of 120 mL of nonionic iodinated contrast material (350 mg I/mL iohexol [Omnipaque, Nycomed Amersham]) at 2 mL/sec. Scanning was performed approximately 1 min after injection using the same parameters as those described for unenhanced scanning.
Images were reviewed with soft-tissue window settings (center, 10 H; width, 410 H) and bone window settings (center, 1,750 H; width, 170 H).
MRI Examinations
MRI examinations were performed on a 1.5-T system (Signa LX, GE Healthcare)
using a phased-array coil. All scans consisted of at least T1- and
fat-suppressed T2-weighted images in two planes. Transverse conventional
spin-echo T1-weighted and fast spin-echo T2-weighted images were obtained
using 5- to 10-mm section thickness, 1- to 2-mm intersection spacing, 2-4
excitations, 512 x 192 matrix, and a variable field of view depending on
the extremity. For T1-weighted spin-echo images, the TR range/TE range was
450-500/10-20. For T2-weighted fast spin-echo images, the parameters were a
TR/TE of 3,000/75 with an echo-train length of 8-16. Sagittal or coronal (or
both) conventional spin-echo T1 and STIR images were also obtained. STIR
images were obtained with a TR/TE range of 3,000/40-50 and inversion time of
150-160 msec.
In 12 cases, unenhanced and dynamic contrast-enhanced T1-weighted fast multiplanar spoiled gradient-echo images with fat saturation (TR/TE, 180/2.1; flip angle, 90°) were obtained with 5- to 10-mm section thickness, 1- to 2-mm intersection spacing, a 512 x 192 matrix, and variable field of view depending on the extremity. Two to three postcontrast sets of images were obtained at 1- to 2-min intervals after IV administration of 0.1 mmol/kg gadopentetate dimeglumine (Magnevist, Berlex Laboratories) at 2 mL/sec.
Reviewers and Procedures
Two reviewers retrospectively reviewed the radiography, CT, and MRI
examinations and determined the presence or absence of findings independently.
Subsequently, discrepant cases were reexamined by both reviewers together, and
a consensus opinion was reached regarding the findings. Reviewers had
knowledge of the presence of fractures, but they had no knowledge of age,
clinical history, or potential underlying tumor.
All images were viewed as hard-copy films. Reviewers had access to all available imaging and were under no fixed time restrictions.
Radiography, CT, and MRI examinations were observed in random order and viewed independently during several separate viewing sessions (as convenient for the reviewer) over approximately 3 weeks to minimize bias due to reviewer order effects and fatigue. After an independent review, imaging was reviewed again by both reviewers to obtain their consensus opinion approximately 4 weeks later.
For radiographs, the reviewers were asked to rate the presence of a pathologic fracture on a 1-3 scale (< 50% sure, 50-95% sure, > 95% sure, respectively). They recorded which bone was involved and the site of fracture (epiphysis, metaphysis, diaphysis) to confirm that these fractures correlated with those seen on CT or MRI. Criteria used for determining the presence of a pathologic fracture included a discrete well-defined, ill-defined, moth-eaten, or permeative underlying bone marrow pattern of destruction; aggressive periosteal reaction; endosteal scalloping; mineralized matrix; and a soft-tissue mass. Criteria used for determining the presence of a stress fracture included a fracture line with callus formation and the absence of features of a pathologic fracture.
For CT scans, the reviewers were asked to rate the presence of a pathologic fracture on a 1-3 scale (< 50% sure, 50-95% sure, > 95% sure, respectively). They recorded the site of fracture (epiphysis, metaphysis, diaphysis) and whether it was comminuted. For pathologic fractures, the reviewers then recorded whether they thought a discrete underlying lesion was present in the marrow. They were asked to characterize the lesion according to one of the five following categories: well-defined with sclerotic margin, well-defined without sclerotic margin, ill-defined, moth-eaten, or permeative [3]. The density of the marrow was recorded as less dense, isodense, or more dense compared with fatty marrow. The presence of endosteal scalloping, a soft-tissue mass, mineralized matrix, and periosteal reaction was recorded. The reviewers recorded if the periosteal reaction was benign (thick, uninterrupted) or malignant (sunburst appearance, hair-on-end appearance, Codman triangle, onion peel appearance) [3]. Reviewers recorded whether IV contrast material was administered and whether there was linear or nodular contrast enhancement around the fracture site. Finally, they recorded whether other lesions were visible on the images.
For MR images, the reviewers were asked to rate the presence of a pathologic fracture on a 1-3 scale (< 50% sure, 50-95% sure, > 95% sure, respectively). They recorded the site of fracture (epiphysis, metaphysis, diaphysis) and whether it was comminuted. They recorded whether a bone marrow signal abnormality was present on T1-weighted imaging; if it was hypointense, isointense, or hyperintense compared with normal muscle; and if the abnormality was well-defined or ill-defined. Similarly, reviewers recorded whether a bone marrow signal abnormality was present on T2-weighted imaging; if it was hypointense, isointense, or hyperintense compared with hematopoietic marrow; and if the signal was well-defined or ill-defined. The presence of endosteal scalloping, periosteal signal abnormality, a soft-tissue mass, and signal voids was recorded. Reviewers were asked to determine whether there was an intracortical signal abnormality around the fracture site and if it was intermediate or high signal on T1- and T2-weighted imaging. The degree of circumferential cortical involvement was determined (< 25%, 25-50%, 50-75%, > 75%). The reviewers recorded the presence of abnormal signal in the adjacent muscle; if it was hypointense, isointense, or hyperintense compared with normal muscle on T1-weighted imaging; and if it was intermediate or hyperintense on T2-weighted imaging. They also characterized the muscle signal abnormality as feathery or as nodular and well-defined. The reviewers recorded whether IV contrast was administered and if there was linear or nodular contrast enhancement around the fracture site. The presence of other visible lesions was also recorded. Finally, the presence or absence of a "fluid" sign, a sign described by Baur et al. [25] that aids in the distinction of vertebral pathologic and benign fractures, was recorded.
Analysis
For the first hypothesis—Cross-sectional imaging aids the
radiologist in detecting pathologic fractures with greater certainty than
radiography—a receiver operating characteristic (ROC) curve analysis was
performed for each reviewer's determination of whether a pathologic fracture
was present. Interobserver agreement was also determined, and kappa scores
were computed. In addition, using the consensus interpretation of whether an
underlying lesion was present at the fracture site, we compared the overall
accuracy of CT for cases that had dedicated extremity imaging with that for
cases that had a routine chest or abdominal CT examination in an effort to
determine the effect of differences in the CT protocols.
For the second hypothesis—Features exist on CT and MRI that distinguish pathologic fractures from stress fracture)—a chi-square analysis or Fisher's exact test was used to compare the CT and MRI features that were determined by consensus to distinguish stress fractures from pathologic fractures. The sensitivity, specificity, positive predictive value, and negative predictive value of each of the MRI and CT features were calculated.
For the third hypothesis—MRI is superior to CT for the identification of pathologic fractures given its increased sensitivity to bone marrow abnormalities—the sensitivity and specificity of CT and MRI were computed in an overall fashion on the basis of whether the consensus interpretation determined an underlying lesion to be present at the fracture site. Differences were considered statistically significant at a p value of less than 0.05.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
) based on determination of a
pathologic fracture was highest for MRI (0.81 for radiography, 0.73 for CT,
and 0.90 for MRI). Interobserver agreement () based on confidence
scores was also highest for MRI (0.59 for radiography, 0.47 for CT, and 0.74
for MRI).
|
Overall CT accuracy was increased for studies performed as dedicated extremity examinations (85%) compared with limb examinations included as part of routine chest or abdominal CT examination (80%).
For the second hypothesis, (imaging features on CT and MRI), Tables 2 and 3 list the CT and MRI features of stress fractures and pathologic fractures. For CT, significant differences included increased marrow density, endosteal scalloping and malignant periosteal reaction, and signs of pathologic fractures. For MRI, significant differences included a well-defined hypointense T1 bone marrow signal abnormality, well-defined hyperintense T2 bone marrow signal abnormality, endosteal scalloping, the presence of a soft-tissue mass, and nonspecific presence of muscle signal abnormalities with pathologic fractures compared with stress fractures. The presence of other lesions was significantly more common with pathologic fractures than with stress fractures. Cortical and muscle T1 and T2 signal abnormalities did not differ significantly. The fluid sign was present in 33% (8/24) of pathologic fractures and 24% (5/21) of stress fractures (p = 0.7). The CT and MRI features of stress fractures and pathologic fractures are shown in Figures 1A, 1B, 1C, 1D, 2A, 2B, 2C, 2D, 2E, 3A, 3B, 4A, 4B, 4C, and 5.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The location of stress fractures and pathologic fractures was different (p < 0.0005). Of 32 pathologic fractures in 30 patients, 3% (1/32) were epiphyseal, 75% (24/32) were metaphyseal, and 22% (7/32) were diaphyseal; of 29 stress fractures, 17% (5/29) were epiphyseal, 24% (7/29) were metaphyseal, and 59% (17/29) were diaphyseal.
For the third hypothesis (CT vs MRI), the overall sensitivity, specificity, positive predictive value, and negative predictive value based on whether consensus opinion of the reviewers revealed a pathologic fracture with an underlying lesion were 96%, 100%, 96%, and 96% for MRI and 84%, 83%, 91%, and 71% for CT, respectively.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Our first hypothesis that cross-sectional imaging aids the radiologist in detecting pathologic fractures with greater certainty than radiography was not entirely substantiated. ROC curve analysis for both reviewers showed that radiography remains an excellent test for distinguishing stress fractures from pathologic fractures, although for both reviewers, MRI had the greatest diagnostic accuracy and confidence of diagnosis for the detection of a pathologic fracture (accuracy of radiography vs MRI, 94% vs 98% for reviewer 1 and 88% vs 93% for reviewer 2) (Figs. 2A, 2B, 2C, 2D, and 2E). However, CT, surprisingly, had lower accuracy than the other two techniques. The difficulty that reviewers had in confidently identifying an underlying lesion at the site of a pathologic fracture on CT may be explained by the overlapping CT features observed for stress fractures and pathologic fractures in this study (Figs. 4A, 4B, and 4C). Low interobserver agreement between the reviewers for CT features further supports the notion that underlying bone marrow lesions in a pathologic fracture may be difficult to confidently detect using CT. Furthermore, the lack of 3D CT and lack of dedicated extremity imaging in some cases likely in part contributed to the overall lower CT accuracy in characterizing fractures (80% accuracy observed with nondedicated studies vs 85% for dedicated studies). It has been shown that axial CT may be inadequate in showing pathology; 3D CT can provide useful additional information for diagnosis over axial imaging alone [27, 28].
Nevertheless, although the absence of 3D CT with some cases is a significant limitation of this study, identification of some sensitive and specific CT features was sufficient in this study. In our patients, CT features of stress fractures generally agree with those previously reported in the literature, although many CT features were common to both stress fractures and pathologic fractures. The role of CT in the diagnosis of a stress fracture has been well established [7, 11-15, 20]. The typical appearance of a stress fracture on CT is that of focal callus formation and endosteal thickening around a fracture site with occasional increased medullary cavity density and adjacent soft-tissue swelling [11, 12, 20].
The most specific features that we detected for the presence of a pathologic fracture were endosteal scalloping, an aggressive periosteal reaction, and hyperdense bone marrow. Unfortunately, in two of 12 cases of stress fracture, reviewers incorrectly interpreted the presence of an underlying lesion, affirming that the CT appearance of a stress fracture may be aggressive and misleading, as predicted by previous reports of the appearance of stress fractures on radiography [3, 7, 8]. In four of 25 cases of pathologic fractures, reviewers missed the presence of an underlying tumor at the site of fracture; they asserted that insufficient specific features for pathologic fracture were identifiable (Figs. 4A, 4B, and 4C). With the advent of 16-MDCT and associated isotropic data sets, it is possible that the increased fine detail of bones afforded by this advanced technology may provide additional specificity for the diagnosis of these fractures.
Unlike the results for CT, the overall performance of MRI in detecting a pathologic fracture was excellent with a sensitivity and specificity of 96% and 100%, respectively. The appearance of a stress fracture on MRI has been described already [13, 16-19, 21]. According to Lee and Yao [21], MRI findings are discernable before radiographic abnormalities and features include decreased marrow signal on T1-weighted imaging and increased marrow signal on T2-weighted imaging. Signal changes in the bone marrow around stress fractures are rather nonspecific and seemingly similar to pathologic fractures. However, in our series, the most useful discriminating feature was that of a well-defined low-signal T1-weighted abnormality, which had a high sensitivity (100%) and specificity (93%) for pathologic fracture. This latter feature is echoed in a report by Yuh et al. [29] indicating that complete replacement of signal within a vertebral body on T1-weighted imaging is a distinguishing characteristic of malignant vertebral fractures compared with benign fractures.
Our hypothesis that MRI was superior to CT for the detection of an underlying bone marrow lesion was confirmed. Furthermore, the negative predictive values of the absence of T1- and T2-weighted signal abnormalities were 100% each, showing that no changes in the bone marrow on MRI rule out an underlying lesion (Figs. 3A and 3B).
The significance of T2 signal abnormalities around a fracture site has been discussed [16, 25]. Increased T2 signal in a subchondral benign femoral fracture corresponds to the proliferation of fibroblasts, chronic inflammation, and fluid exudates in viable bone [16]. Baur et al. [25], on the other hand, described the fluid sign, a sensitive sign for the presence of a benign vertebral fracture, which corresponds to osteonecrosis at the fracture site. In our study, the fluid sign was not proven to be a useful feature in distinguishing long bone fractures. Thus, in stress fractures, T2 signal changes may indicate inflammation; in pathologic fractures, T2 signal changes are likely a mixture of tumor and inflammatory changes. The assessment of T1 signal changes is therefore paramount.
Other MRI features that were significantly different between stress fractures and pathologic fractures were the presence of a soft-tissue mass and endosteal scalloping, features common to CT. Periosteal signal changes and cortical signal changes were not helpful differentiating features and were better characterized on CT.
Compared with stress fractures, pathologic fractures are more commonly associated with muscle signal abnormalities. These abnormalities, however, were not significantly different in character (most were feathery in nature) and likely simply represent edema rather than tumor infiltration into the surrounding tissues. Hanna et al. [30] reported that massive edema in the muscles surrounding a bone tumor was an ominous clinical finding, more commonly found in malignant rather than benign underlying lesions and typically involving the disruption of a muscle attachment to bone by the tumor.
Finally, according to our series, the location of the fracture may provide a clue to the type of fracture given that 75% of pathologic fractures are metaphyseal, whereas 58% of stress fractures are diaphyseal. Our results concur with those reported in the literature; fatigue fractures in one study were all located in the diaphysis [19].
Limitations
Regarding case selection in this study, all patients who underwent
radiography also went on to have cross-sectional imaging by CT and MRI. Thus,
presumably, features observed on radiography and the clinical circumstances of
these patients were inconclusive regarding the cause of the patients' symptoms
[31]. We have no information
on patients with stress fractures or pathologic fractures who were treated
without cross-sectional imaging. The accuracy of CT and MRI may be diminished
in this study relative to that in the general population if cross-sectional
imaging is performed only in clinically difficult cases or cases with
inconclusive radiographic findings. On the other hand, it is also possible
that cross-sectional imaging is performed in cases in which clinical signs and
symptoms are obvious; the latter circumstance would result in inflated
accuracy relative to that in the general population.
We studied fractures that were biopsy-proven or with typical expected clinical outcome to identify distinguishing features of these entities. Such study criteria introduce workup bias because it is likely that imaging had an influence on whether these patients underwent biopsy. The true agreement between imaging and biopsy results may be less than that observed in this study if every patient with a fracture in the general population underwent biopsy. However, the latter scenario is clinically inappropriate and avoidance of workup bias in this study is impossible.
Only cases of known fractures were included in the study, and reviewers were aware of the presence of a fracture. This may be unrealistic because, on occasion, patients are imaged at a later phase of healing when a fracture line may not be visible after bone remodeling. In one study, the fracture line in tibial longitudinal stress fractures was seen in only 82% and 73% of the fractures on MRI and CT, respectively [13]. In our series, in some cases, the fracture line was not very well depicted on radiography (Figs. 2A, 2B, 2C, 2D, and 2E), but reviewers still scored these cases with regard to the presence of a pathologic fracture.
The variety of pathology included was a deliberate attempt to simulate real-world interpretation of lesions, but is a limitation of the study in that different entities are likely to produce varied imaging appearances.
Although reviewers did not have access to clinical history, they were able to glean information from the images themselves to bias their interpretation of a particular fracture site. For example, observation of an open physis allowed the determination of age, which may have guided their interpretation. However, as already stated, it should be emphasized that the clinical setting of a fracture may be deceptive and may lead to erroneous treatment. A young athletic patient who may be susceptible to a fatigue fracture is also in an age group in which insidious tumors may be present, given that almost 1% of pediatric patients with musculoskeletal complaints have an underlying malignancy [32]. Conversely, insufficiency fractures, which occur more commonly in oncology patients and in the elderly, are often unsuspected. Skeletal scintigraphy and PET scans ordered in these patients may show activity at the site of a stress fracture [33], and only careful attention to imaging features will distinguish a stress fracture from a metastatic lesion or pathologic fracture.
Regarding the technical aspects of the CT studies performed in this series, all examinations were performed with an MDCT scanner. However, with the advent of isotropic data sets and 3D CT, bone detail may be further enhanced to more easily detect the fracture lines of a stress fracture. The subtle destruction of the cortex and bone marrow of a pathologic fracture also may become more visible and characterization of the underlying bone marrow lesion more obvious than with older CT technology. Similarly, MRI examinations consisted of T1-weighted, T2-weighted, and gadolinium-enhanced sequences. Advanced techniques that may be helpful in differentiating benign from malignant fractures were not performed in this study and include diffusion-weighted imaging and chemical shift imaging [22, 23, 34, 35].
Our results regarding the utility of cross-sectional imaging over radiography in the evaluation of fractures would likely be altered in the axial skeleton; for example, in the sacrum, radiographs are notoriously insensitive to disease and both CT and MRI would be expected to provide much greater sensitivity and specificity in the evaluation of fractures at that site.
Finally, we should note that the results of this study can be applied only to patients who present with a fracture; they cannot necessarily be generalized to patients with a bone lesion who have not sustained a fracture.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
