| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Emergency Radiology |
1 From the Department of Radiology, University of California-Davis Medical Center, 4701 X St, Sacramento, CA 95817 (R.W.K., P.F.B., M.I., C.S.B., W.R.N., R.A.M., F.K.O., D.M.B., V.C.P., B.W.C.) and the Department of Biostatistics, University of California at Davis (C.M.D., R.A.L.). Received October 15, 1998; revision requested November 11; revision received December 28; accepted April 30, 1999. Supported in part by Hitachi, Tokyo, Japan. Address reprint requests to R.W.K.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Conventional radiography and MR imaging were performed in 199 patients presenting to a level 1 trauma center with suspected cervical spine injury. Weighted sensitivities and specificities were calculated, and a weighted average across eight vertebral levels from C1 to T1 was formed. Fourteen parameters indicative of acute injury were tabulated.
RESULTS: Fifty-eight patients had 172 acute cervical injuries. MR imaging depicted 136 (79%) acute abnormalities and conventional radiography depicted 39 (23%). For assessment of acute fractures, MR images (weighted average sensitivity, 43%; CI: 21%, 66%) were comparable to conventional radiographs (weighted average sensitivity, 48%; CI: 30%, 65%). MR imaging was superior to conventional radiography in the evaluation of pre- or paravertebral hemorrhage or edema, anterior or posterior longitudinal ligament injury, traumatic disk herniation, cord edema, and cord compression. Cord injuries were associated with cervical spine spondylosis (P < .05), acute fracture (P < .001), and canal stenosis (P < .001).
CONCLUSION: MR imaging is more accurate than radiography in the detection of a wide spectrum of neck injuries, and further study is warranted of its potential effect on medical decision making, clinical outcome, and cost-effectiveness.
Index terms: Magnetic resonance (MR), comparative studies, 31.121411, 31.121412 • Spinal cord, injuries, 341.41, 341.42, 341.444 • Spine, CT, 31.1211 • Spine, injuries, 31.41, 34.42, 34.444 • Spine, MR, 31.121411, 31.121412 • Spine, radiography, 31.11 • Trauma, 31.40
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Magnetic resonance (MR) imaging, with use of systems with either high or low field strength, is of increasing utility in the evaluation of acute injuries to the cervical spine (5–14). This is in spite of the fact that MR imaging long has been perceived as being incompatible with the unpredictable clinical status and intensive monitoring requirements for patients in the emergency department. However, MR imaging provides unparalleled multiplanar depiction of soft-tissue injury, including ligament damage, intervertebral disk herniation, spinal cord injury, and prevertebral and paravertebral hemorrhage or edema after blunt cervical spine injury. Because MR imaging is not routinely used in most trauma centers, its utility as a frontline diagnostic modality is not settled. We installed a dedicated MR imaging trauma unit immediately adjacent to the emergency department in May 1994. Herein, we describe our findings in evaluation of the detection of cervical spine injury with conventional radiography compared with open-design, middle field strength MR imaging. Our study was prospective and was performed prior to spinal injury reduction. The readers were blinded, and we used a detailed statistical model whereby each vertebral level from C1 to T1 was assessed separately as a unit of observation.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
In the statistical analysis, the data were pooled from both arms of the study patients. All 199 patients underwent a comprehensive neurologic examination. Most patient trauma was the result of motor vehicle accidents (n = 98 [49%]), followed by falls (n = 30 [15%]), assault (n = 22 [11%]), bicycle and motorcycle accidents (n = 20 [10%]), and automobile versus pedestrian accidents (n = 10 [5%]).
All patients underwent a comprehensive conventional radiographic assessment with the following projections: lateral, lateral swimmer's, anteroposterior, opened-mouth odontoid, Fuch odontoid, and both posterior oblique projections. The quality of images was assessed on-line by a radiology resident or attending staff member. Additional radiographs were acquired as deemed necessary, and in some cases digital images were acquired.
All patients were examined with use of an open-design 0.3-T MR imaging unit (MRP 7000; Hitachi, Tokyo, Japan). Unlike the standard 1.5-T high field strength MR imaging systems, this open-design, vertical, middle field strength MR imaging system allows direct patient visualization and the use of standard ventilatory and infusion devices in addition to MR imaging–compatible monitoring equipment. Imaging was performed with a minimum of three sequences by using a solenoid neck coil: sagittal T1 weighted (650/25 [repetition time msec/echo time msec]),T2-weighted gradient echo (560/35, 20° flip angle), and fast spin echo (4,500/117). Section thickness was 4 mm with a 1-mm gap. If any abnormalities were identified, additional axial T1-weighted (760/25) and gradient-echo (500/35, 20° flip angle) images were obtained. The total imaging time for a three-sequence examination was 20 minutes. A five-sequence study required approximately 35 minutes. A gradient-echo sequence with a high flip angle (250/15, 70° flip angle) was sometimes used to improve contrast between the intervertebral disk and the cerebrospinal fluid.
Computed tomography (CT) (model 900; Toshiba, Kawasaki, Japan) was performed in 17 of the 199 patients in the study. At our institution, CT is used only to further assess the possibility or to further clarify the extent of osseous abnormalities. We therefore performed CT in a relatively small number of cases, and the comparison of CT scans with MR images and conventional radiographs is thus only descriptive. The CT protocol consisted of 5- or 3-mm contiguous axial sections of the local area of concern, and images were obtained without the administration of intravenous or intrathecal of contrast agents. Sagittally reformatted images were also obtained routinely.
With use of each vertebral level as the unit of observation, an anatomic location was defined as the portion of the closest vertebra that intersected a horizontal line drawn through the abnormality. Locations were thus established and tabulated for the nearest vertebral level. If any component of a lesion extended to a contiguous anatomic segment, then it also was tabulated to involve that segment. Fourteen parameters indicative of acute injury were assessed and tabulated for each of the eight spinal levels encompassing the first cervical (C1) to the first thoracic (T1) vertebrae as follows: acute fracture, subluxation or dislocation of the facets, subluxation or dislocation of the vertebra, traumatic disk herniation, epidural hemorrhage, cord edema, cord hemorrhage, cord compression, cord transection, anterior longitudinal ligament damage, posterior longitudinal ligament damage, interspinal ligament damage, prevertebral hemorrhage or edema, and paravertebral hemorrhage or edema. In addition, preexisting disease of the cervical spine was assessed and included healed vertebral fracture, disk spondylosis, facet spondylosis, end-plate marrow change, focal disk herniation, foraminal stenosis, canal stenosis, and degenerative subluxation. Thus, there was a tabulation of 22 anatomic characteristics at eight separate levels (ie, C1–T1) in 199 patients for both conventional radiography and MR imaging, which equates to 70,048 distinct observations.
Because surgical confirmation was rarely available and surgical observations are not thought to be completely accurate (8), we established a standard of truth at each vertebral level. This standard of truth was determined by two experienced radiologists (P.F.B., V.C.P.) who worked in consultation and had access to all available clinical and imaging information, which included conventional radiographs, MR images, and CT scans obtained at the time of presentation in the emergency department or subsequently. All information available from the clinical records of the emergency department, prehospitalization records, inpatient records with discharge summary, and any surgical records were also used. Review was not performed until all of these elements were collated. During this review, observations were recorded on the data sheets, which were identical to those used by the blinded readers. If any disagreements arose between the two reference standard readers, a consensus was forced at that time. There were no persistent or major disagreements during this detailed and rigorous review process. Some subsequent MR images used by the reference standard readers were acquired on a 1.5-T MR system (GE Medical Systems, Milwaukee, Wis).
A separate two-person set of readers (C.S.B., F.K.O.) worked completely independently to assess the conventional radiographs at each vertebral level, and a second two-person set of readers (W.R.N., B.W.C.) also worked independently to assess the MR imaging findings at each vertebral level. The conventional radiograph and MR image readers were blinded to any clinical information or any of the other imaging findings. The readers did not consult, and their evaluations were tabulated separately. The CT scans were assessed separately by the same two neuroradiologists, who each assessed the MR images at a different time and also without access to the clinical information.
Many of the criteria used to characterize injury as depicted at MR imaging were reported by Flanders et al (10). Fracture is determined on the basis of alteration in the configuration of the vertebra or break in the cortical continuity with or without alteration in the signal intensity on T1- and T2-weighted images. In comparison to the surrounding marrow, edema of the marrow of the vertebra is depicted as a region with low signal intensity on T1-weighted images and as a region with higher signal intensity on T2-weighted images. Spinal cord edema is determined on the basis of the presence of intramedullary foci of high signal intensity on T2-weighted images. Intramedullary hematoma is determined on the basis of the presence of foci of high signal intensity on T1-weighted images or foci of low signal intensity on T2-weighted, spin-echo, or gradient-echo images. Traumatic disk herniation is defined as disk protrusions associated with increased signal intensity of the disk tissue on T2-weighted images or as the presence of associated injuries to the paraspinal soft tissues or spinal cord at the same level. Ligament injury is defined as high signal intensity in the ligament itself or frank interruption of the dark band of the ligaments on T2-weighted images. Pre- or paravertebral hemorrhage or edema is defined as regions of low signal intensity on T1-weighted images and as regions of high signal intensity on T2-weighted images, with or without the presence of soft-tissue widening or swelling.
The presence of preexistent spondylosis or stenosis on either MR images or conventional radiographs was determined by means of subjective assessment. For example, disk space narrowing, marginal osteophytes, and configurational changes of vertebrae with sclerosis were findings considered degenerative in nature. Fracture at conventional radiography was defined as a break, either complete or incomplete, in the cortical continuity of a bone, with or without alteration in configuration. Dislocation of vertebrae was determined as a complete disruption of the normal contact between articular surfaces. Subluxation was defined as the circumstance in which the joint surfaces maintain some degree of contact but without normal alignment. Soft-tissue edema or hemorrhage on conventional radiographs was determined if there was widening of the retropharyngeal space, widening of the retrotracheal space, displacement of the prevertebral fat stripe, or tracheal displacement in either the anteroposterior or lateral projections. Ligament injury was suspected with widening of the anterior or posterior intravertebral spaces or widening of the interspinal space.
Sensitivity and specificity were the basis for statistical comparisons between MR imaging and conventional radiography. A standard approach for calculating these measurements is the recording of injury status of the patient as detected with the modality and the comparison with a reference standard that is assumed to be a measure of injury status without error. This method, however, neglects the potential for disagreement at each vertebral level. A theoretic combination of possibilities is shown in Figure 2. Note, for example, in Figure 2 that a patient could have fracture in six hypothetical scenarios, but only scenario A represents complete agreement of the modality with the reference standard. Observations within the same subject do not necessarily provide the same information as observations in different patients. We calculated weighted average sensitivities and specificities at each vertebra and formed a weighted average across the eight levels. CIs were constructed based on these weighted averages and accounted for correlations in vertebrae in the same subject.
|
2 test was used, with the patient as the unit of observation, to assess associations between injury types (such as cord edema) and risk factors (such as cord stenosis) with significance accepted when the P value was less than .05. Reader agreement for conventional radiography and MR imaging was measured by means of 
statistics (15). |
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
2), acute skeletal fracture (P < .001), and canal stenosis (P < .001). There was no association between degenerative spinal disease and ligament injury (P = .726). There were 21 vertebral fractures in 14 patients, 18 acute facet subluxations or dislocations in nine patients, and 26 acute vertebral subluxations or dislocations in 13 patients. Superficial soft-tissue injuries such as prevertebral hemorrhage or edema were found in 37 patients at 142 levels, and paravertebral hemorrhage or edema was found in 32 patients at 141 levels. There was anterior longitudinal ligament damage in 13 patients involving 26 levels, posterior longitudinal ligament damage in eight patients at 16 levels, and interspinal ligament damage in six patients at 18 vertebral levels. Traumatic disk herniation was noted in 10 patients at 21 levels, cord edema was noted in 15 patients at 39 levels, and cord compression involved 15 patients at 36 levels. Epidural hemorrhage was noted in only one subject, and there were no cases of cord hemorrhage or cord transection. Chronic changes detected in this patient population are shown in Table 2. Degenerative subluxation, end-plate marrow changes, and facet spondylosis are noted to occur in an older patient population with a mean age of approximately 64 years.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
, 0.4), for cord edema of 42% (CI: 22%, 64%; , 0.6), and for cord compression of 38% (CI: 18%, 59%; , 0.7). The weighted average sensitivities and CIs for MR imaging versus conventional radiography are displayed for acute injuries in Table 3 and for chronic changes in Table 4. There was no significant difference in the 43% weighted average sensitivity of MR imaging (CI: 21%, 66%; , 0.5) versus the 48% weighted average sensitivity for conventional radiography (CI: 30%, 65%; , 0.6) for detection of acute fracture. In the 14 patients with fractures, these two modalities depicted fractures in nine of the same patients and in five different patients (three at conventional radiography and two at MR imaging). When positive findings at MR imaging and conventional radiography are taken together, there was only one complete miss in 14 patients (93% detection rate).
|
|
, 0.6), paravertebral hemorrhage or edema (weighted average sensitivity, 60%; CI: 42%, 77%; , 0.6), and anterior longitudinal ligament damage (weighted average sensitivity, 46%; CI: 23%, 70%; , 0.4) (Table 3). MR imaging showed a tendency to be better than conventional radiography for depicting injuries to the posterior longitudinal and interspinal ligaments, though there was a slight overlap in CIs. The statistical assessment of MR imaging versus conventional radiography for chronic changes is displayed in Table 4. MR imaging was significantly better for the evaluation of canal stenosis. MR imaging showed a weighted average sensitivity of 37% (CI: 19%, 57%) for focal disk herniation. For the detection of degenerative subluxation, facet spondylosis, and foraminal stenosis, conventional radiography was statistically superior to MR imaging.
As was done for the assessment of weighted average sensitivities at each vertebral level, we performed similar analysis on the specificities of each imaging modality; however, a lack of false-negative findings—large weighted average specificities, all close to 99%—did not allow a significant separation.
We were unable to acquire enough CT data to construct CIs, so the results are descriptive. The CT evaluation of bone injuries on a per patient unit of observation showed a weighted average sensitivity of 86%. When assessed on a per vertebra level, the weighted average sensitivity was 65%. In the 10 patients with fractures who underwent CT assessment, there were two misses, for an 80% detection rate versus a 93% detection rate in the 14 patients with MR images plus conventional radiographs. In the 17 patients examined by means of CT, there were 72 injuries. In this subset of patients, conventional radiographs depicted 18 (25%) of these findings, MR images depicted 62 (86%), and CT scans depicted 15 (21%). Thus, CT did not seem to be competitive with MR imaging for the depiction of acute injuries other than fractures, but these results cannot be proved owing to the small sample size.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
We found equal weighted average sensitivities for fracture detection in comparing radiography and MR imaging. With all modalities available, however, the reference standard readers recognized that fracture detection with MR imaging required a substantial period of learning. With careful analysis and re-review, most osseous fractures could be seen on MR images by the reference standard readers. Fracture displacement, bone marrow edema, and adjacent soft-tissue edema or hemorrhage were substantial aids in fracture detection at MR imaging. The need for training in interpretation cannot be overemphasized.
Flanders et al (10) and others (11–14) have established that MR imaging should be considered the standard of reference in the detection of soft-tissue injury associated with cervical spine trauma. In their study, the sensitivity of MR imaging in the depiction of disk protrusion was 44%, of spinal cord swelling was 33%, of spinal cord compression was 77%, and of prevertebral edema was 35%. On the other hand, conventional radiographic and CT findings have been suggested by Goldberg et al (7) to have little correlation with the patient's neurologic status and thus to be of little or no prognostic value. Kulkarni et al (6) used MR imaging at 1.5 T and have suggested a high degree of correlation between three cord injury patterns and patient outcome. The first pattern described is edema due to only contusion, which indicates a more favorable prognosis with potential for reversibility. The second pattern is that of blood alone, which indicates a more substantial axonal injury and much less potential for reversibility. The third pattern is a mixed pattern of blood and edema, which indicates an intermediate prognosis with some potential for recovery (16). The craniocaudal length of the segment of abnormal cord signal intensity from contusion has been correlated with prognosis by Schaefer et al (17). Their study results indicate that if cord contusion is greater than one vertebral level in length, the prognosis is less favorable.
Emergency physicians routinely assess clinically for evidence of instability once radiographs are declared negative. Instability can result from fracture alone, ligament disruption alone, or a combination of these injuries. As defined by White and Southwick (18), clinical instability is the inability to maintain normal associations between vertebrae while under a physiologic load. Instability may lead to subsequent injury of the spinal cord or nerve roots or substantial pain or deformity. Presence of instability or neurologic deficit should prompt consideration of an MR imaging assessment. If substantial pain is present or if instability is suggested on radiographs, further evaluation with flexion-extension radiography is sometimes performed. The reliability of this examination in the acute phase of injury has been questioned (19). Furthermore, the complete cervical spine, including the cervicothoracic junction, is not routinely visible on lateral flexion-extension radiographs, and radiographs can only imply instability by demonstrating abnormal motion. MR imaging, on the other hand, allows direct visualization of not only morphologic changes within the bones and soft tissues, including the ligaments, but also of signal intensity abnormality usually due to edema or blood. The multiplanar capabilities of MR imaging allow excellent assessments of the anterior and posterior longitudinal and flaval ligaments, which are the three major ligament groups believed to provide stability.
Posttraumatic disk herniation is demonstrated exceptionally well at MR imaging and was depicted in 10 patients at 21 vertebral levels, which was much less prevalent than the findings reported by Flanders et al (10). None of these disk herniations were depicted on either conventional radiographs (Table 3) or CT scans in our series. Secondary findings of cervical disk damage include increased signal intensity of the injured disk on the T2-weighted images, which possibly represents edematous changes and the presence of associated injuries to the paraspinal soft tissues or spinal cord at the same level. The presence of a herniated disk fragment does not necessarily imply a deleterious clinical consequence (10).
The relation between preexistent cervical spine skeletal abnormalities—such as spondylosis, central canal stenosis, and acute fracture—and cord or ligament injury is controversial (20). We found a significant association between the occurrence of cord injury in conditions of underlying spondylosis, central canal stenosis, and acute vertebral fracture.
Patients with spondylosis who have hyperextension injury are at higher risk for central cord syndrome (21). The mechanism leading to the central cord injury is posterior disk bulging and anterior buckling of the hypertrophied flaval ligament with hyperextension. The central gray matter and adjacent tracts are injured, which accounts for the predominance of weakness in the upper extremities and the eventual decrease in pain and temperature sensation. These injuries may be radiographically occult (Figs 3–5) (22). MR imaging, however, can demonstrate injury to the spinal cord and ligament injury or disk injury, such as a fracture through the disk or disk extrusion.
Our study in many ways parallels that of Orrison et al (23), in which MR imaging was used prospectively in acute spinal trauma. Our rationale is similar to the one used in their study, and we used a middle field strength MR imaging system that allows examination of patients who are critically ill. The combination of open design and middle to low field strength enables both direct access and the ability to monitor patients who are difficult to examine in a more traditional high field strength MR imager. The latter have limitations of confined space and strong magnetic fields, which make it difficult to provide respiratory and cardiovascular monitoring, traction devices, and accessibility to the patient for suctioning and direct clinical observation. In a retrospective review of 113 consecutive spinal trauma cases in which a low field strength (0.064 T) magnet was used, Orrison et al (23) determined that MR imaging was superior to either CT or conventional radiography in the evaluation of soft-tissue or ligament injury. They found, however, that MR imaging had a substantially lower rate of positive findings for fracture than did CT, and it was also significantly less sensitive for fracture than was conventional radiography. Spinal cord contusion, epidural hematoma, high-grade stenosis, and ligament or soft-tissue injury were best evaluated with MR imaging. Our results differ somewhat in that MR images obtained at 0.3 T were not significantly different from conventional radiographs in the depiction of fracture and appeared to have a complementary role. We agree, however, as has also been suggested by Orrison et al (11), that MR imaging may be the modality of choice in acute cervical spine trauma and that CT should be reserved for selected patients with complex fractures.
Problematic with any study such as this is the lack of what might be considered a definitive standard of truth with which comparisons can be made. As mentioned by Flanders et al (10), surgical confirmation of MR imaging findings is not regularly available nor are the observations made at the time of surgery thought to be completely accurate because anterior surgical decompression does not allow direct visualization of the damaged disk in situ (8,10). Even in that important investigation (10), the standard of truth in the comparison of CT with MR imaging, the modality reputedly offering the best definition of a particular abnormality, was used as the reference standard. Thus, CT was considered the reference standard in all observations that related to disruption of the bone axis, and MR imaging was the reference standard for all observations related to the spinal cord and paraspinal soft tissues. This method has been the rule in the literature rather than the exception in assessment of the relative merits of different modalities in cervical spine trauma. We have attempted to establish a more detailed and comprehensive standard of truth by having two experienced radiologists reviewing all conventional radiographs, all MR images at the time of presentation or thereafter, all CT scans, the emergency department clinical record, the prehospital clinical record, inpatient records with the discharge summary, and surgical records if surgery was performed. The final determination of the standard of truth was completed after all the latter elements were collated and assessed.
The value of a prospective study such as ours is that it allows the possibility of gaining insights into the relative merits of differing modalities without being heavily biased toward patients with injury. The disadvantage of our approach, however, is the relatively low prevalence of traumatic disease, which led us to be more selective in the second group of patients (n = 98). For example, in the first group of consecutive patients (n = 101), acute fractures were demonstrated in only 3% (three patients), cord edema in only 4% (four patients), and cord compression in only 5% (five patients). We were more selective in the second group of patients (n = 98), and acute fractures were demonstrated in 11% (11 patients), cord edema in 11% (11 patients), and cord compression in 10% (10 patients). In spite of this, our selection bias is not nearly as great as that in previous retrospective studies such as that by Kulkarni et al (6), in which 70% (19 of 27) of patients had cord abnormalities and 78% (21 of 27) had skeletal or ligament injuries. In the study by Mirvis et al (8), all 21 patients had acute neurologic deficits following cervical spine trauma, and in the study by Flanders et al (10) all 78 patients had a demonstrable cervical fracture or subluxation or a neurologic deficit. Thus, the second arm of our investigation, though tending toward a higher prevalence of traumatic disease, is not nearly as strongly weighted toward severe injury as has been reported in the more common retrospective studies, which allows a greater depth of appreciation about the true sensitivity of a modality for any abnormality. Also advantageous to us was our ability to perform imaging prior to treatment.
Another weakness of this investigation was the relatively limited number of CT imaging assessments in this patient population. CT examinations were performed in only 8% (17 of 199) of patients. Thus, our comparison of CT with conventional radiography and MR imaging can be only anecdotal. Our use of CT in trauma to the cervical spine is, however, not necessarily unusual. In a 19-month study of 4,134 patients admitted to a shock trauma center, cervical spine CT examinations were requested in only 9.9% (408 of 4,135) of patients (4). We hope other prospective studies, perhaps similar to ours, can incorporate a more comprehensive analysis of the relative merits of CT, conventional radiography, and MR imaging in the initial presentation of patients with acute neck injuries.
In conclusion, MR imaging provides an opportunity for unparalleled detection of soft-tissue injury, including ligament damage, intervertebral disk herniation, spinal cord injury, and pre- and paravertebral hemorrhage or edema after cervical spine injury, yet it is rarely used in most trauma centers. The results of this prospective study of 199 patients confirm the clinical feasibility and utility of middle field strength MR imaging in the initial assessment of cervical spine trauma. We believe that MR imaging should be strongly considered in the early evaluation of cervical spine injury.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
3 Current address: Aurora, Ohio.
4 Current address: Department of Radiology, University of Utah School of Medicine, Salt Lake City, Utah.
Author contributions: Guarantors of integrity of entire study, R.W.K., P.F.B.; study concepts, R.W.K., P.F.B., C.M.D., M.I., R.A.L., V.C.P., B.W.C.; study design, R.W.K., P.F.B., C.M.D., R.A.L., C.S.B.; definition of intellectual content, R.W.K., P.F.B., C.M.D., M.I., R.A.L.; literature research, R.W.K., P.F.B., C.S.B.; clinical studies, R.W.K., P.F.B., C.S.B., W.R.N., R.A.M., F.K.O., D.M.B., V.C.P., B.W.C.; experimental studies, R.W.K., P.F.B., C.M.D., M.I., R.A.L.; data acquisition, all authors; data analysis, R.W.K., C.M.D., M.I., R.A.L., D.M.B.; statistical analysis, C.M.D., M.I., R.A.L.; manuscript preparation, R.W.K., P.F.B., C.M.D., R.A.L., C.S.B.; manuscript editing, R.W.K., P.F.B., C.S.B.; manuscript review, R.W.K., P.F.B., C.M.D., R.A.L., C.S.B., W.R.N., R.A.M., F.K.O.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Harrison and C. Cairns Clearing the cervical spine in the unconscious patient CEACCP, August 1, 2008; 8(4): 117 - 120. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Shanmuganathan, R.P. Gullapalli, J. Zhuo, and S.E. Mirvis Diffusion Tensor MR Imaging in Cervical Spine Trauma AJNR Am. J. Neuroradiol., April 1, 2008; 29(4): 655 - 659. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Wee, J. H Reynolds, and A. Bleetman Imaging after trauma to the neck BMJ, January 19, 2008; 336(7636): 154 - 157. [Full Text] [PDF]
|
|
|
|
|
|
D. Goradia, K.F. Linnau, W.A. Cohen, S. Mirza, D.K. Hallam, and C.C. Blackmore Correlation of MR Imaging Findings with Intraoperative Findings after Cervical Spine Trauma AJNR Am. J. Neuroradiol., February 1, 2007; 28(2): 209 - 215. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. J. Hogan, S. E. Mirvis, K. Shanmuganathan, and T. M. Scalea Exclusion of Unstable Cervical Spine Injury in Obtunded Patients with Blunt Trauma: Is MR Imaging Needed when Multi-Detector Row CT Findings Are Normal? Radiology, October 1, 2005; 237(1): 106 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. W. Sliker, S. E. Mirvis, and K. Shanmuganathan Assessing Cervical Spine Stability in Obtunded Blunt Trauma Patients: Review of Medical Literature Radiology, March 1, 2005; 234(3): 733 - 739. [Abstract] [Full Text] [PDF]
|
|
|
|
