| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Emergency Radiology |
1 From the Institute of Radiology (G.R., S.M.), Department of Orthopaedic and Trauma Surgery (G.M.), and Center for Clinical Research (D.S.), Unfallkrankenhaus Berlin Trauma Center, Warener Str 7, 12683 Berlin, Germany; and Institute of Radiology, Ernst-Moritz-Arndt University of Greifswald, Greifswald, Germany (N.H.). Received December 25, 2004; revision requested February 28, 2005; revision received March 18; accepted April 15. Address correspondence to D.S. (e-mail: dirk.stengel{at}ukb.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: The institutional review board approved this study; informed consent was not required. Charts and images of consecutive patients treated for multiple trauma (injury severity score, >16) between January 1998 and October 2003 were reread by an experienced radiologist. Until October 2002, subjects were screened for BCVI with US. Since November 2002, patients underwent CT angiography of the carotid and vertebral arteries. Sensitivity and specificity of US and CT angiography were calculated with 95% confidence intervals (CIs).
RESULTS: The early cohort included 1471 patients (mean age, 35.8 years ± 17.7 [standard deviation]), and the late cohort included 407 patients (mean age, 39.2 years ± 18.8). US depicted five blunt vessel injuries but later missed another eight, which led to cerebral ischemia. With a BCVI frequency of 0.9%, sensitivity and specificity of US were 38.5% (95% CI: 13.9%, 68.4%) and 100% (lower 95% confidence limit, 99.7%), respectively. In the second cohort, the BCVI rate was 2.7%. CT angiography depicted BCVI in 11 patients, with a sensitivity of 100% (lower 95% confidence limit, 71.5%), but produced one false-positive result.
CONCLUSION: Injuries to the cervical arteries among blunt trauma patients are more common than previously reported. Duplex Doppler US has inadequate sensitivity to help rule out this condition. The notable morbidity with missed dissections warrants routine contrast material–enhanced studies of the carotid and vertebral vessels if patients are scheduled for CT of the cervical spine.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In multiple trauma, the primary survey gives priority to injuries requiring immediate intervention because of their detrimental effect on circulation and gas exchange or profuse bleeding (eg, pericardial tamponade, tension pneumothorax, or unstable pelvic fractures). Life-threatening blunt injuries to the neck and cervical spine are rare. Temporary stabilization with a rigid collar permits emergency physicians to focus on diagnostic work-up of the body cavities and to return to the neck region after excluding substantial torso trauma.
Established diagnostic algorithms call for conventional radiography to help detect cervical spine fractures. However, the extracranial vessels are not regularly addressed with these protocols. Several authors (3,4) suggested clinical signs and risk factors (eg, basilar skull and cervical spine fractures, bruises, and expanding hematoma) that may increase the prior probability of BCVI and should prompt further imaging with use of digital subtraction angiography (DSA).
Recently, Cothren and colleagues (5) reported their results with this staged approach. Of 13 280 subjects admitted during a 6-year period, 643 (4.8%) met the criteria for further imaging and underwent DSA. Of those, 114 had confirmed carotid injury. The authors estimated an overall BCVI prevalence of 0.86%. One might argue with the chosen denominator, since the article does not contain information about the 12 637 subjects not undergoing early angiographic imaging.
Application of diagnostic reference tests to those patients who have obvious signs of neck injury is straightforward but does not meet the principles of screening. According to the Suburban Emergency Management Project of Benedictine University, Lisle, Ill (available at: http://www.ben.edu/semp/index.html), screening can be defined as "the presumptive identification of unrecognized disease or defect by the application of tests, examinations or other procedures, which can be applied rapidly. Screening tests sort out apparently well persons who probably have a disease from those who probably do not."
It is unclear whether the absence of clinical predictors is helpful to exclude BCVI. Neurologic examination at admission is unreliable in ventilated patients and may not reveal signs of impaired cerebral blood flow in the early stages of vascular occlusion. The situation gets worse if patients had encountered injuries that will ultimately require surgery (eg, evacuation of intracranial hematoma, splenorrhaphy, or splenectomy), which will delay the time until a definite diagnosis on the extracranial vessels can be made.
The purposes of this study were to retrospectively evaluate the frequency of BCVI in patients with multiple trauma and to retrospectively compare the accuracy of duplex Doppler ultrasonography (US) and computed tomographic (CT) angiography by using clinical follow-up and subsequent imaging as reference standards.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
We recorded information independent of regular hospital documentation on a four-page registration form for the Trauma Registry of the German Association of Trauma Surgeons (6). A board-certified surgeon (G.M.) and a board-certified surgeon and clinical epidemiologist (D.S.) were responsible for clinical data collection throughout the study period. The database included names of all patients admitted to our emergency department with the clinical diagnosis of multiple trauma. Eligible patients had encountered a high-velocity injury (eg, a fall from a height or a vehicle crash leading to death of another vehicle occupant). The diagnosis of multiple trauma was made in cases with an injury severity score of greater than 16, either suspected by emergency physicians on the scene or by early survey at the trauma bay (7).
Data collection included patient demographics, injury characteristics, physiologic conditions at admission, health status before and after surgery, and clinical course during critical care. Patients were followed up until day 90 after the accident. For this study, three authors (G.R., G.M., and D.S.) surveyed the electronic hospital charts for further clinical information, including coding of diagnoses, procedures, and complications.
The institutional review board approved the retrospective investigation of data sets compiled during trauma work-up along locally established standards of care. Informed consent was not required. We were requested (a) to limit our analysis on available data (ie, no attempts were made to contact patients again for study purposes) and (b) to ensure anonymity by complying with the national laws for data protection and safety.
Diagnostic Work-up for Severe Trauma
At our hospital, whole-body contrast material–enhanced CT is the primary imaging tool after clinical examination, focused abdominal sonography for trauma, placement of central lines, and fluid resuscitation.
Patients suspected of having multiple trauma are regularly scheduled to undergo CT, regardless of clinical signs and symptoms or negative results from focused abdominal sonography for trauma. At our hospital, the CT suite is located close to the trauma bay, which allows for a rapid transfer and short examination times. Invasive monitoring is maintained during imaging, and anesthesiologists may intervene immediately if patients show signs of hemodynamic deterioration. Hypotensive patients are excluded from this protocol in cases of refractory hypotension and positive results from focused abdominal sonography for trauma, which require immediate laparotomy, or US signs of pericardial tamponade and/or active thoracic bleeding, which demand emergency thoracotomy.
The radiology department provides a 24-hour in-house service. The board-certified radiologist on duty performs and interprets all imaging procedures in real time. Local quality maintenance and improvement measures comprise the following steps: (a) initial diagnosis is made by the radiologist and/or surgeon on duty; (b) conflicting results are resolved by consensus to account for both clinical signs and symptoms and radiologic findings; and (c) images are presented in daily grand rounds held in the morning and in the afternoon.
The original scanning protocol was developed for a Tomoscan AVE (Philips Medical, Eindhoven, the Netherlands) CT scanner (8). Sequential CT of the skull and helical CT from the cervical spine to the upper thoracic spine without contrast enhancement were performed. Scanning parameters were 3-mm section width, 5 mm/sec table feed, 3-mm reconstruction index, 120-kVp tube voltage, and 250-mAs tube current. We administered a 150 mL bolus of nonionic contrast agent (iohexol, Omnipaque; Schering, Berlin, Germany) by means of a power injector (Medrad Medical Systems, Volkach, Germany) to scan the trunk from the upper thoracic aperture to the femoral lesser trochanter. The maximum scan volume involved 99 rotations, with a rotation time of 1 second. The contrast agent flow rate was 2.5 mL/sec; scanning started after a 30-second delay. We created secondary multiplanar reconstructions in sagittal and coronal planes. In October 2002, the Tomoscan AVE was replaced with a multi–detector row CT scanner (Somatom Sensation 4; Siemens, Erlangen, Germany). Clinical management, documentation, and follow-up schemes remained unchanged.
A board-certified consultant radiologist (G.R.) with 6 years of experience in all imaging techniques relevant to the trauma setting reviewed all archived digital CT angiograms, US scans, and further imaging studies obtained during the study period by using a picture archiving and communication system workstation (Easy Access; Philips Medical, Eindhoven, the Netherlands).
Screening for BCVI: Early Cohort (January 1998 to October 2002)
The single-section scanning protocol did not allow for a simultaneous performance of CT angiography to screen for BCVI. Thus, after completion of CT and before transfer to the critical care unit, all patients underwent duplex Doppler US of the extracranial vessels irrespective of clinical signs of cervical injury (eg, hematoma of the neck, seat belt marks). Patients brought directly to the operating theater underwent cervical US within 24 hours.
Since 1998, six board-certified radiologists have been part of the staff responsible for in-house service, including night shifts. All of them manage the entire spectrum of imaging procedures, including duplex Doppler US, CT, CT angiography, MR imaging, MR angiography, and DSA. Since 2000, two extra board-certified radiologists (including G.R.) have been admitted to this team. Before independent US examinations and interpretation of all other types of images were allowed, the team members had to be trained for at least 2 years. The head of the department (S.M.) is an instructor of the German Society for Ultrasound in Medicine and has more than 15 years of experience with diagnostic US.
We assessed flow in the common carotid artery, external carotid artery, proximal internal carotid artery, and vertebral artery on both sides. For all examinations, a Sonoline Elegra (Siemens) US scanner with a 7.5- or a 5.0-MHz linear-array transducer, where suitable, was used.
We recorded hypoechoic intramural hematoma, dissected arterial walls, luminal narrowing, hemodynamically relevant stenoses, or occlusions as pathologic findings indicating blunt arterial injury. Confirmation procedures consisted of CT angiography, MR angiography, or DSA, depending on the condition of the patient. Subjects received heparin, clopidogrel, and aspirin at the discretion of the critical care team and in accordance with planned surgical procedures.
Screening for BCVI: Late Cohort (November 2002 to October 2003)
The new hardware offered the opportunity of simultaneous performance of CT angiography of the extracranial and basilar vessels, while examining the cervical spine and the skull base. The patient was placed on the table with the arms remaining at the sides. To adjust for varying cardiac output and to yield optimal contrast in the supraaortal vessels, an 80-mL bolus of a nonionic contrast agent (iohexol) was administered with a power injector at a flow rate of 4 mL/sec after semiautomated bolus tracking in the ascending aorta, with a threshold of 90 HU. A second bolus of 60 mL of iohexol was immediately administered with a flow rate of 3 mL/sec for the imaging of thoracic and abdominal vessels, as well as parenchymal organs.
Scanning parameters included four detector rows with 2.5-mm section thickness, 7 mm/sec table feed, 120-kVp tube voltage, and 250-mAs tube current. Reconstructions overlapped 1.25–3.0 mm of section thickness. After triggering, CT angiographic data acquisition lasted 10 seconds. Similarly to the first protocol, we obtained multiplanar images in sagittal and coronal planes.
Pathologic patterns suggesting dissection were irregular vessel walls, tapering stenoses, occlusions, or bulges consistent with a pseudoaneurysm. Treatment was initiated similarly to that in the early cohort.
Reference Standards
The independent reference test applied to all patients was clinical follow-up until day 90. Participants were followed up as outpatients conjointly by general practitioners and ambulatory surgeons.
Subjects with inadequate weaning from respiratory support, with altered mental state, suspected of having stroke or intracranial bleeding, or with signs of ischemia on scheduled cerebral CT scans after head injury or neurosurgery underwent further imaging procedures. Definite imaging tests comprised DSA and MR angiography.
Any dissection missed at early examination and proved with further imaging studies represented a false-negative result. This also applied to outpatients who developed neurologic sequelae. We considered any blunt artery dissection suspected at primary survey but that remained unproven with further studies a false-positive finding. An uneventful clinical course without a documented neurologic disorder related to ischemic events represented a true-negative result.
Statistical Analysis
To ensure correct classification of a disease, we sampled US and CT angiographic scans that indicated the case status (ie, proved BCVI) and paired these images in a 1:2 fashion with a random sample of true-negative findings obtained during both study intervals. The images were independently evaluated (G.R. and S.M.) for presence or absence of BCVI. Interrater agreement was measured with 
statistics.
We estimated measures of diagnostic accuracy (ie, sensitivity, specificity, likelihood ratios, and prior and posterior probabilities) to compare the duplex Doppler approach with the CT angiographic protocol in disclosing dissection.
We also determined the risk ratio of arterial damage with certain accompanying injuries. We calculated 95% confidence intervals (CIs) of point estimates. We explored differences in age between the groups by using the t test for independent samples and assessed differences in proportions with 
2 statistics.
To gain a first impression of the efficiency of either method (ie, the effect on patient-centered outcomes), we calculated the differences in proportions and the number needed to treat for neurologic deficits and mortality. The 95% CI was estimated by using the hybrid-score method (9).
All analyses were performed by using software (STATA, version 8.0; Stata, College Station, Tex).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
|
|
|
Altogether, subjects with cervical spine injuries had a threefold increased relative risk of blunt carotid artery dissection (risk ratio, 3.47; 95% CI: 2.44, 4.93). However, with a specificity of 83% (1522 of 1830) and a sensitivity of 58% (14 of 24), the absence of spine injuries was an almost useless criterion to rule out BCVI. The predictive features of head injuries and midfacial fractures were only slightly better, with sensitivity of 67% (16 of 24) and 71% (17 of 27) and specificity of 37% (677 of 1830) and 65% (1189 of 1830), respectively. values indicated substantial to almost perfect agreement between independent raters (11) (Table 2).
|
|
|
|
|
|
Eight patients attracted attention during follow-up because of neurologic sequelae indicative of cerebral ischemia (ie, ptosis, anisocoria, hemiparesis, and syncope). The interval between the index test and the onset of symptoms ranged from 1 to 40 days (median, 11 days). Definite imaging techniques depicted infarctions caused by four one-sided dissections of the internal carotid artery, one bilateral dissection, one occlusion, and one dissection of the vertebral artery (Figs 3, 4). One of these patients died of cerebral contusion and epidural bleeding. The remaining patients recovered, with moderate to severe disability. Three of 13 patients were referred to the vascular surgery service for carotid endarterectomy with venous patch.
|
|
|
|
US depicted three of seven BCVIs (four of which were carotid dissections) accompanying cervical spine fractures. By confining the analysis to this high-risk population with a total of 270 subjects, sensitivity increased slightly to 42.9% (95% CI: 9.9%, 81.6%), whereas specificity remained unchanged.
Accuracy of CT Angiography
CT angiography depicted BCVI in 11 patients in the late cohort, with a frequency of 2.7% (95% CI: 1.4%, 4.8%). MR angiography later helped verify all pathologic findings. Three patients also underwent DSA before planned interventional therapy or surgery. CT angiography produced one inconclusive result classified as a false-positive finding. MR angiography helped exclude blunt injury to the extracranial vessels.
Injury patterns comprised five one-sided and one bilateral carotid dissection, as well as four one-sided and two bilateral dissections of the vertebral artery (Fig 5). All carotid dissections involved the arterial segment passing the carotid canal. We noted six false aneurysms and three complete occlusions.
|
|
Sensitivity and specificity of CT angiography were 100% (lower 95% confidence limit, 71.5%) and 99.1% (95% CI: 98.6% 100%), respectively. Four patients died within 24 hours. No autopsy was requested, and we again considered these cases inconclusive. If all these patients had vascular injury, sensitivity decreased to 73.3% (95% CI: 44.9%, 92.2%).
US and CT Angiography
Figure 5 depicts the expected diagnostic value with use of duplex Doppler and CT angiographic screening for varying prevalences in fictitious populations. Since positive findings provided with either method are decisive, we calculated posterior probabilities only for negative findings. Given a prevalence of blunt carotid dissection of 1%, a negative duplex Doppler scan slightly decreased the residual chance to 0.61% by a negative likelihood ratio of 0.62.
A negative CT angiogram helped rule out arterial dissection (with a negative likelihood ratio close to null). By accounting for the lower 95% confidence limit of sensitivity (ie, 71.5%), a negative CT angiogram yielded at least a negative likelihood ratio of 0.29 and a posterior probability of 0.28%.
By using the actual numbers of patients with BCVI (13 at US and 11 at CT angiography) and the related numbers of adverse events (eg, eight moderate and two severe deficits), it can be shown that the regular use of CT angiography and the interventions chosen as a consequence of CT angiographic findings produce better neurologic outcomes. The risk difference for moderate to severe neurologic deficits in favor of early CT was 43.4% (95% CI: 4.0%, 67.9%), and the number needed to treat was 2.3 (95% CI: 1.5, 24.9).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Although not explicitly proved under the rigor of a randomized controlled trial, anticoagulation has established itself in clinical practice. Current best evidence supporting the application of heparin in BCVI patients comes from the Cothren et al study (5), which formally met level IIb evidence (prospective cohort study with >80% follow-up) (available at: http://www.cebm.net/levels_of_evidence.asp). One might even discuss whether the results meet level Ic evidence ("all-or-none"), since none of the 73 patients with BCVI receiving heparin had a stroke, whereas 19 of 41 patients with contraindication to anticoagulation had an ischemic event. This leads to a risk difference of –46% in favor of anticoagulation, with the 95% CI ranging from –62% to –31%. In other words, providing rather than withholding anticoagulants will prevent one extra event in roughly any second patient with BCVI. Given this strong size effect—despite the possible bias inherent to the observational design—the equipoise principle is violated, making a placebo-controlled trial almost impossible.
A population-based prevalence below 3% would demand exceptionally high sensitivity of the chosen screening modality. Besides high sensitivity, it must be available at the point of care to avoid unnecessary delay during primary survey.
Duplex Doppler US has many features of a promising screening tool for BCVI: rapidity, mobility, cost-effectiveness, and noninvasiveness. Like focused abdominal sonography for trauma, it can be performed simultaneously in resuscitation, clearing of other torso injuries, and temporary fixation of fractures. In a systematic review and meta-analysis of the available literature, Nederkoorn et al (13) calculated pooled sensitivities of 86% and 96%, respectively, for the US detection of 70% stenosis and 99% occlusion, compared with DSA studies; MR angiography had pooled sensitivities of 95% and 98%, respectively. However, the review did not specifically address the issue of blunt trauma to the extracranial vessels.
US failed to depict eight of 13 intimal tears in our study. Possible causes for this unsatisfying performance include inadequate surrounding conditions, time pressure, and hampering owing to stiff necks and central venous lines inserted in the jugular vein. In addition, US has technical limits in assessing the distal parts of the carotid artery close to the siphon and skull base, as well as the vertebral arteries.
In contrast, in our sample CT angiography depicted all blunt vessel injuries at admission. During follow-up, there was no evidence of missed injuries that harmed cerebral perfusion. Our data underline the diagnostic accuracy of CT angiography and support its use as a screening tool to disclose BCVI in patients with blunt multiple trauma (14,15).
Multi–detector row CT easily allows for the acquisition of an angiogram without conflicting rapid trauma work-up. The median latent interval until onset of neurologic symptoms before routine CT angiography was 11 days, and one might argue that each day on anticoagulants contributes to rescue of the brain. Cerebral infarction is often diagnosed accidentally during or shortly after weaning. Regardless of the initially normal cerebral CT scan, tissue loss is irreversible when recognized.
We did not specifically address the efficiency of screening for BCVI (ie, actions taken because of a test result and the related clinical benefit). We stress this was a retrospective study, and the noted outcome benefits for patients undergoing CT angiography must be interpreted with caution. According to our data, one extra moderate to severe neurologic deficit may be prevented in any second patient by using primary CT angiography rather than US-based algorithms to exclude BCVI. This perfectly matches the results published by Cothren et al (5). We believe this difference substantiates the role of early CT angiography in managing blunt multiple trauma, although one might speculate whether patients with an early diagnosis of BCVI would have progressed to cerebral ischemia without anticoagulants or would have remained clinically silent anyway. Again, our analysis was exploratory, not confirmative, and requires proof with a large multicenter randomized controlled trial.
Limitations of our study merit further discussion. First, it was a retrospective chart review. However, all subjects undergoing index tests during both observation periods were followed up according to the policy of a national trauma registry. We believe that these precautions reduced residual misclassification of disease. Nevertheless, we cannot exclude some underreporting of transient neurologic deficits that should have prompted further imaging.
Second, it is obviously difficult to compare the diagnostic accuracy of duplex Doppler US and CT angiography by using two different cohorts of patients. However, the study hospital serves as the major regional trauma center, and referral practices did not change since 1998. Also, recruitment periods were manageable; sample sizes were large; and entry criteria, documentation procedures, and follow-up rules were not affected by replacing US with CT angiography. Thus, it is reasonable to assume that the observed effects were induced by the choice of diagnostic modality and not by an imbalance in baseline profiles.
Helical single-section CT did not allow simultaneous performance of CT angiography of the extracranial vessels, and we considered duplex Doppler US the best adjunctive method to gain a first idea of the presence or absence of blunt carotid injury. A reliable comparison would have required to continue with US after the introduction of the multi–detector row scanner and CT angiography. However, we did not want to offset the time gained by the more rapid CT imaging without a reasonable chance of collecting valuable added information with a second imaging test.
Third, one might argue that the radiologist (G.R.) who had already reviewed the charts and images should not have been involved in the independent rating procedure. We accepted this possible methodological weakness for logistic and capacity reasons. Recall bias, which is unlikely because of the large sample size, may influence indices of diagnostic accuracy but is of minor importance as far as interrater agreement is concerned. Statistics help assess whether two independent observers consent to findings beyond chance, regardless of the correctness of their appraisal. Theoretically, both could fail to correctly interpret any test result but nevertheless show almost perfect interrater agreement.
Finally, the design was still susceptible to partial verification bias, since only positive findings of imaging tests or later clinical symptoms led to further diagnostic work-up. There might be an unknown prevalence of minor hemodynamically ineffective intimal tears missed with both index tests. This chance was obviously higher in the early cohort.
It is difficult to select a proper diagnostic reference standard given the accuracy of modern CT angiography. In a recent meta-analysis, Koelemay et al (16) noted an 85% sensitivity with CT angiography in cases of atherosclerotic carotid stenosis of 70%–99%; for carotid occlusion, the sensitivity was estimated at 97%. These data apply to chronic alterations of the vessel wall and must therefore be interpreted with caution. However, they represent the current best evidence to deal with the value of CT angiography in disclosing pathologic conditions of the extracranial arteries. The results generate, at least, robust prior probability of CT angiographic accuracy, and it is unlikely that CT angiography performs much worse in the trauma setting.
Clinical follow-up may only help identify patients with symptomatic BCVI. However, its independent application (ie, irrespective of negative or positive findings of the index test) is a logistically and ethically acceptable alternative to the general verification of the diagnosis by means of invasive reference standards such as DSA. For ethical reasons, we did not mandate another reference test such as DSA or MR angiography in cases of negative findings. The sensitivity of CT angiography might have been overestimated but is likely to settle between the confidence limits.
Although we have not performed a formal cost-effectiveness analysis, our experience shows that CT costs are a minor contribution to the gross volume spent for trauma care. In a review that appeared in Radiology, Novelline et al (17) emphasized potential cost cuts that could be achieved by using CT (eg, to clear injuries to the cervical spine) instead of conventional studies such as radiography with additional oblique projections.
In conclusion, our data suggest a higher rate of BCVI in patients with multiple trauma than has been found in previous studies. Further research must confirm these estimates. Duplex Doppler US has inadequate sensitivity to exclude intimal tears. On the basis of our findings, we recommend contrast-enhanced CT of carotid and vertebral arteries in all trauma patients scheduled to undergo CT of the cervical spine. A negative high-quality CT angiographic result, while not definitive proof of normal cervical vessels, is highly reliable in helping predict a benign clinical course.
|
FOOTNOTES |
|---|
Abbreviations: BCVI = blunt cerebrovascular injury • DSA = digital subtraction angiography • CI = confidence interval
Authors stated no financial relationship to disclose.
Author contributions: Guarantor of integrity of entire study, D.S.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, G.M., D.S.; clinical studies, S.M., G.R., G.M., D.S.; statistical analysis, D.S.; and manuscript editing, S.M., G.R., N.H., D.S.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Langner, S. Fleck, M. Kirsch, M. Petrik, and N. Hosten Whole-Body CT Trauma Imaging with Adapted and Optimized CT Angiography of the Craniocervical Vessels: Do We Need an Extra Screening Examination? AJNR Am. J. Neuroradiol., November 1, 2008; 29(10): 1902 - 1907. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. W. Sliker Blunt Cerebrovascular Injuries: Imaging with Multidetector CT Angiography RadioGraphics, October 1, 2008; 28(6): 1689 - 1708. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Rodallec, V. Marteau, S. Gerber, L. Desmottes, and M. Zins Craniocervical Arterial Dissection: Spectrum of Imaging Findings and Differential Diagnosis RadioGraphics, October 1, 2008; 28(6): 1711 - 1728. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. W. Sliker, K. Shanmuganathan, and S. E. Mirvis Diagnosis of Blunt Cerebrovascular Injuries with 16-MDCT: Accuracy of Whole-Body MDCT Compared with Neck MDCT Angiography Am. J. Roentgenol., March 1, 2008; 190(3): 790 - 799. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
