| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Depts of Neurosurgery and Neuroradiology, Institute of Neurological Sciences, Southern General Hosp, Glasgow, Scotland (P.M.W.); and the Depts of Clinical Neurosciences (P.M.W., J.M.W., V.E.) and Medical Statistics (V.E.), Univ of Edinburgh, Bramwell Dott Building, Western General Hosp, Crewe Rd, Edinburgh EH4 2XU, United Kingdom. Received Dec 1, 1999; revision requested Jan 19, 2000; revision received Mar 20; accepted Mar 24. P.M.W. and V.E. supported by the British Brain and Spine Foundation from the Davie Cooper Scottish Aneurysm Study grant, administered by the Univ of Glasgow. J.M.W. supported by the Medical Research Council under the Clinical Research Initiative in Clinical Neuroscience. Address correspondence to P.M.W. (e-mail: pmw@skull.dcn.ed.ac.uk).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: A 1988–1998 literature search for studies with 10 or more subjects in which noninvasive imaging was compared with angiography was undertaken. Studies meeting initial criteria were evaluated by using intrinsically weighted standardized assessment to determine suitability for inclusion. Studies scoring greater than 50% were included.
RESULTS: Of 103 studies that met initial criteria, 38 scored greater than 50%. CT angiography and MR angiography had accuracies per aneurysm of 89% (95% CI: 87%, 91%) and 90% (95% CI: 87%, 92%), respectively. For US, data were scanty and accuracy was lower, although the CIs overlapped those of CT angiography and MR angiography. Sensitivity was greater for detection of aneurysms larger than 3 mm than for detection of aneurysms 3 mm or smaller—for CT angiography, 96% (95% CI: 94%, 98%) versus 61% (95% CI: 51%, 70%), and for MR angiography, 94% (95% CI: 90%, 97%) versus 38% (95% CI: 25%, 53%). Diagnostic accuracy was similar for anterior and posterior circulation aneurysms.
CONCLUSION: CT angiography and MR angiography depicted aneurysms with an accuracy of about 90%. Most studies were performed in populations with high aneurysm prevalence, which may have introduced bias toward noninvasive examinations. Supplemental material.
Index terms: Aneurysm, intracranial, 17.73 • Computed tomography (CT), angiography, 17.12116 • Digital subtraction angiography, comparative studies, 17.12483 • Magnetic resonance (MR), vascular studies, 17.12142 • Ultrasound (US), Doppler studies, 17.12983, 17.12984
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The aim of this study was to perform a systematic review of the evidence on the diagnostic accuracy of each of the three noninvasive methods compared with that of intraarterial DSA to determine whether accuracy was influenced by aneurysm prevalence, sample size, or aneurysm site or size in the populations studied. Our purpose was also to determine whether the information already available is sufficient to be confident in the accuracy of noninvasive examinations or more information is needed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
If studies met the initial criteria, they were then formally assessed for eligibility by two authors (P.M.W., J.M.W.) by using predetermined quality criteria with a standardized assessment form. Each reviewer independently assessed each study (ie, two reviews per study). This checklist method enabled an objective and reproducible assessment of each article to be performed. The form contained a checklist of 26–27 items relevant to studies of diagnostic performance. The items were grouped into three main categories, which are summarized in Table 1. The predetermined weights applied to each item by the observers in deriving an assessment of each article also are given in Table 1. The actual outcome of a study in terms of diagnostic performance—that is, how well the noninvasive examination performed—was not assessed with regard to inclusion or exclusion in the meta-analysis.
|
statistic for interrater agreement was 0.84 (95% CI: 0.73, 0.95), which was indicative of very good agreement.
Search Strategy
Following advice from the Cochrane Database of Systematic Reviews Stroke Review Group Search coordinator (B. Thomas, personal communication, October 1997), one author (P.M.W.) searched MEDLINE and EMBASE electronic databases for relevant articles by using exploded headings under the terms "tomography," "x-ray computed," "magnetic resonance angiography," and "ultrasonography" from January 1988 to December 1998. In addition, searching was performed by using free-text terms, including all possible variations of "CT angiography," "MR angiography," and "transcranial ultrasound." Both strategies were combined by using the Boolean operator, "OR." This search was then combined with a second search under "subarachnoid hemorrhage" and "intracranial aneurysms" (again derived from both exploded headings and free-text terms) by using the Boolean operator, "AND." The search then was narrowed to studies with adult humans.
For all the studies identified by using this search, the two authors (P.M.W., J.M.W.) cross checked the reference lists for additional articles and checked the reference lists of relevant review articles. This method of cross checking was continued until no further studies were identified. One of the authors (P.M.W.) also hand searched the journals not indexed in either of the above electronic databases in which relevant articles were identified from reference lists. The authors did not specifically search meeting abstracts, because only articles published in full could meet our inclusion criteria; however, they did check for subsequent publications of potentially relevant works that were originally presented as an abstract.
The search strategy was validated by performing a hand search (P.M.W.) of the RSNA Index to Imaging Literature, published by the Radiological Society of North America, and the journals Neurosurgery, Journal of Neurosurgery, and Stroke, which are the three most common journals quoted in the reference lists that were not covered in the RSNA Index during 1988-1997. The RSNA Index includes references to 42 major peer-reviewed journals on all diagnostic imaging modalities but with an emphasis on current cross-sectional imaging modalities. By using the hand search of the RSNA Index, 32 articles that met the initial inclusion criteria were identified; 31 (97%) of these were picked by using the electronic search. The hand search of Neurosurgery, Stroke, and the Journal of Neurosurgery yielded 11 articles that met the initial inclusion criteria. All of these were identified by using the electronic search.
Data Extraction
Two authors (P.M.W., J.M.W.) independently extracted data from the identified studies by means of the standardized data extraction form (ie, two reviews per study). For non–English-language publications, data were extracted with the aid of a translator with a medical science background. The data extracted from each article are summarized in Table 1. By using the extracted data, the prevalence and distribution of intracranial aneurysms at intraarterial DSA were determined, and from the actual numbers of aneurysms and subjects with a correct or incorrect diagnosis at noninvasive imaging, the true-positive rate for the noninvasive examination versus angiography, true-negative rate, false-positive rate, and negative rates were calculated on both a per-subject and per-aneurysm basis. When the same patients were included in more than one study, we extracted data from only the most recent study to avoid inclusion of the same patients twice. When a second angiogram had been obtained and showed an aneurysm and the first angiogram had been negative, the result of the second angiogram was used for the meta-analysis.
Statistical Data Analyses
Baseline descriptive characteristics were extracted from each study to allow calculation of the mean numbers of patients and aneurysms and of the aneurysm prevalence per study for each modality. Our eligibility criteria permitted inclusion of studies that did not provide results from which data on true- and false-positive rates and negative rates could be extracted per subject and per aneurysm, provided that the study’s overall weighted assessment score was greater than 5, but these studies were excluded from further analysis because the necessary data could not be extracted. By using the data extracted from each study, the true- and false-positive rates and negative rates per aneurysm and per subject were tabulated into 2 x 2 tables for each modality, and the sensitivity, specificity, predictive, and accuracy values were calculated. Exact 95% CIs based on binomial probabilities were calculated (14).
To combine independent studies of the same diagnostic examination for meta-analysis, the method of Moses et al (15) was used. With this method, a logistic transformation of data in simple 2 x 2 tables was used, and then linear regression curves were fit for the true- and false-positive rates by using the least-squares method. The results of these regression curve analyses were used to plot each study into summary receiver operating characteristic curves by using an S-PLUS, version 4.5 statistical software package (STATSCI, Seattle, Wash). By using receiver operating characteristic analysis, which jointly considers the sensitivity and specificity of an examination, greater insight into the difference in performance between examinations is provided compared with that achieved by examining accuracy alone. With the summary receiver operating characteristic method, the area under the curve cannot be determined, but the differences between two examinations can be assessed by comparing the proportion of data points lying above and below the best-fit line with a standard 
2 test. A detailed description of this statistical methodology is given in the article by Moses et al (15).
We also examined the effects of aneurysm prevalence in the study population, study sample size, aneurysm size, recent versus older study, and aneurysm site (ie, anterior vs posterior circulation) on the accuracy of noninvasive modalities in the detection of intracranial aneurysms.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The median sample size for CT angiography was 30 subjects; for MR angiography, 29; and for TCD, 38 (Table 2). The median prevalence of aneurysm in the CT angiography studies was 79.5% (IQR, 74.0–100.0); in the MR angiography studies, 76.0% (IQR, 52.0–97.0); and in the TCD studies, 89.5% (IQR, 79.0–100.0) (Table 2). The median prevalences of aneurysm in the excluded studies were 95.5% (IQR, 67.0–100.0) for CT angiography, 82.5% (IQR, 84.0–100.0) for MR angiography, and 100.0% (IQR, 79.0–100.0) for TCD. The similarities in median averages and IQRs indicated that there was no difference between the excluded and included studies with regard to aneurysm prevalence. In cases in which a second intraarterial DSA study was required—usually owing to technical failure in an elderly or restless patient or to vasospasm—the result of the second angiogram was used. In seven (0.4%) of the 1,765 patients, a second angiogram, obtained after the first image was negative, showed an aneurysm.
|
In 14 (88%) of 16 CT angiography studies, spiral technology was used. In all the MR angiography studies, a three-dimensional time-of flight technique was used. In one study, two-dimensional time-of flight imaging (28) also was used, and in three, time-of flight and phase contrast MR angiography (32,35,42) was used. For TCD, color flow imaging (7,48) was used in two studies and power Doppler US (46,47) was used in two studies. A US contrast agent was used in only one study (47).
Several authors (8,38,42) have prospectively studied the importance of reviewing base as well as reconstructed MR angiograms and concluded that it was very useful. In the current meta-analysis, the material reviewed was explicitly stated in 19 of 20 MR angiography studies. In only three small studies (11,28,39) out of the remaining 19 (including only 79 patients), neither the base nor the reconstructed images were explicitly reviewed. However, the accuracy in this subgroup was very good: Per subject it was 97% (77 of 79), and per aneurysm it was 95% (55 of 58). However, the difference was not substantial compared with the results of the studies in which base images were reviewed, in which the accuracy was 88% (681 of 772) per subject and 89% (699 of 784) per aneurysm, and, owing to the small numbers, the CIs in the subgroup in which neither base nor reconstructed image review was explicit were very wide. In three CT angiography articles, exactly which images were available for review was not explicitly stated; in all the other CT angiography studies, it was indicated that base as well as reconstructed images were available for review, so a similar subgroup analysis was not possible.
Of 16 CT angiography studies, three were performed in patients who were not known to have an aneurysm or recent SAH but had symptoms that could be attributed to an underlying aneurysm (12,15,17); seven were performed in a population in which predominantly all of the patients were known to have an intracranial aneurysm or recent SAH (18–20,23,25,26,43); and six were performed in a population that consisted of a mixture of these groups. The corresponding data for MR angiography studies were three (37–39), 11 (8,27,28,30–32,35,36,40,41,43), and five of 20 studies, respectively. One MR angiography study was performed in an asymptomatic population at increased risk for an aneurysm (4). All TCD studies were performed in patients known to have an aneurysm or recent SAH. Thus, in only one of 38 studies was noninvasive imaging assessed in a group of asymptomatic subjects relevant in screening for intracranial aneurysms.
Results per subject could be extracted from 11 of 16 CT angiography and 18 of 20 MR angiography studies, and results per aneurysm could be extracted from all 16 CT angiography, 18 of 20 MR angiography, and two of four TCD studies. Complete result data—that is, true-positive, false-positive, and negative rates— could be extracted both per subject and per aneurysm from 11 of 16 CT angiography and 16 of 20 MR angiography studies but from none of the TCD studies (ie, 27 of 38 studies in total). The baseline characteristics indicated that the studies of CT angiography and MR angiography were very similar (Table 2). There were much fewer reports on the use of TCD compared with angiography in the detection of aneurysms.
It is important to remember that all study patients who did not undergo corroborative DSA were excluded from further analysis. This means that the number data presented in the meta-analysis results may not match the total number of subjects in a study. For example, in the study by Korogi et al (38), nine of 202 subjects did not undergo DSA, and, therefore, 193 subjects were included in the meta-analysis.
Diagnostic Performance
The full results of the meta-analysis of CT angiography and MR angiography performance on a per-patient basis are presented as summary receiver operating characteristic curves in Figure 1. The results on a per-aneurysm basis can be viewed in an electronic format (Fig E1). The results indicate that both modalities performed well: The curves are concentrated toward the upper left corner of the receiver operating characteristic space, and the scattering of results is close to the fit lines for CT angiography and MR angiography. CT angiography performed marginally better, but the difference was not significant (P = .411). The summary receiver operating characteristic curves on a per-aneurysm basis were very similar, with again no significant difference between CT angiography and MR angiography (P = .09). There were insufficient data to obtain a meaningful summary receiver operating characteristic curve for TCD. A comparison of sensitivity and specificity results, with 95% CIs and meta-analysis results, is presented as Forrest plots for both studies individually in Figure 2. These data clearly illustrate the similarity in sensitivity and specificity between CT angiography and MR angiography.
|
|
|
|
We were unable to analyze performance versus time of the noninvasive examination—for example, before an SAH, in which extensive blood might interfere with the visualization of small aneurysms on both CT angiograms and MR angiograms, because the precise time of the examination relative to the onset of an SAH in most studies was unclear.
The CT angiography studies published before 1995 (two of 16) had an accuracy per subject of 84% (27 of 32) compared with the 14 studies published after 1994, which had an accuracy of 93% (424 of 454); the corresponding accuracy rates per aneurysm were 87% (40 of 46) and 89% (617 of 695), respectively. The MR angiography studies published before 1995 (8 of 20) had an accuracy per subject of 92% (211 of 230) compared with the 12 studies published after 1994, which had an accuracy of 88% (537 of 611); the corresponding rates per aneurysm were 87% (227 of 260) and 91% (527 of 582), respectively. Thus, there was a trend for the more recently performed CT angiography studies, after spiral technology had been well established, to have greater accuracy than the studies performed earlier. This effect was not seen with MR angiography, possibly because of the more mature status of this technique by 1994.
To test for increased bias in the small studies, an analysis of subgroups according to study size was performed. Large was defined as a sample size of 50 or more subjects. (There is evidence that this is the minimum desirable study size for a comparison of diagnostic methods [46]). For CT angiography, there was no substantial difference between the five large and 11 small studies: The accuracy per subject was 94% (583 of 649; 95% CI: 91%, 97%) versus 90% (164 of 182; 95% CI: 85%, 94%), respectively.
For MR angiography per subject, the 14 smaller studies had a substantially higher positive predictive value, 97% (236 of 243; 95% CI: 94%, 99%), and a substantially lower NPV, 66% (51 of 77; 95% CI: 55%, 77%), than did the six larger studies, which had a positive predictive value of 89% (188 of 212; 95% CI: 84%, 93%) and an NPV of 89% (283 of 319; 95% CI: 85%, 92%). Per aneurysm, the larger MR angiography studies had a substantially higher specificity, 97% (218 of 225; 95% CI: 94%, 99%), and NPV, 87% (218 of 250; 95% CI: 82%, 91%), than did the smaller studies, which had a specificity of 83% (35 of 42; 95% CI: 69%, 93%) and an NPV of 45% (35 of 77; 95% CI: 34%, 57%). These data do not support the hypothesis that a preponderance of small studies results in substantial bias that leads to erroneously optimistic results for the diagnostic performance of noninvasive imaging methods (Table E1).
We were able to extract adequate data for an analysis of sensitivity according to aneurysm size from 12 CT angiography and 12 MR angiography studies. In an additional two MR angiography studies, a size breakdown based on factors that were very different from those in the other studies was provided, and in two studies, only a partial breakdown of detection according to size was given. The results of this analysis clearly showed that the sensitivity for aneurysm detection was substantially different between aneurysms with a maximum dimension of 3 mm or less and those larger than 3 mm. A substantial difference in sensitivity was not found between aneurysms 3–10 mm in maximum diameter and those 10 mm or larger (Table 4).
|
A subanalysis of anterior circulation aneurysm detection in different locations was performed. Data on detection by site could be extracted from 11 (69%) of 16 CT angiography and 16 (80%) of 20 MR angiography studies. There were considerable differences between studies with regard to how data were presented for aneurysm site: In some studies, broad categories such as "internal carotid artery" or "middle cerebral artery" were used, whereas in others, more specific data were given. It was possible to extract data on anterior circulation aneurysms in six anatomic sites as follows: anterior communicating arterial aneurysms, "other" anterior cerebral arterial aneurysms, middle cerebral arterial aneurysms, posterior communicating arterial aneurysms, ophthalmic arterial aneurysms, and "other" internal carotid arterial aneurysms. (Three of the six site categories were broad anatomic groupings.) The sensitivity of the imaging techniques for aneurysm detection by location was analyzed independently. Several studies did not include details on the site and/or size of the false-positive aneurysms. Therefore, specificity, NPV, and accuracy could not be calculated.
With CT angiography, sensitivity was poorest for the detection of other anterior cerebral arterial aneurysms—73% (11 of 15; 95% CI: 45%, 92%). The next poorest sensitivity was for the detection of other intracranial aneurysms—88% (50 of 57; 95% CI: 76%, 95%). Because of the relatively small numbers of aneurysms in each site category, the CIs were wide and overlapping. The relatively poor sensitivity for the detection of other anterior cerebral arterial aneurysms was probably because in cases in which the circle of Willis was examined, the aneurysms could have been outside the examined volume. The terminal and intracavernous carotid regions are adjacent to bone, and, thus, aneurysms in these areas, particularly the small ones, are difficult to visualize clearly at CT angiography. With MR angiography, sensitivity was poorest for the detection of posterior communicating arterial aneurysms (41 of 50) and the other intracranial aneurysms (88 of 107), 82%, but again the differences were small. The poorer detection at these sites may have been due to the difficulty in detecting aneurysms, particularly the small ones, at sites with marked vessel tortuosity and overlapping at MR angiography.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The results of direct comparisons between the performances of the noninvasive modalities—for example, CT angiography versus MR angiography—should be interpreted cautiously, because very few patients underwent any two of the noninvasive examinations, which is the most appropriate way to make a true direct comparison. Any apparent differences in modality performance could have been due, at least in part, to differences in the populations studied, in the size and distribution of the aneurysms, and/or in other study variables. Many of the studies included in our systematic review were quite small (even though studies with less than 10 subjects were excluded), and only five CT angiography studies, six MR angiography studies, and one TCD study included 50 or more subjects.
The accuracy of the reference standard is an important factor in any comparison of examinations. We regarded the intraarterial DSA result as definitive, even if it was subsequently proved to be incorrect. This was justified by the rarity with which a second angiogram showed a different result—in seven (0.4%) of 1,765 patients included in the meta-analysis.
It has been reported that noninvasive imaging examinations are much poorer in the detection of aneurysms smaller than 3 mm than in the detection of those larger than 3 mm—for example, a sensitivity of 25% versus 92% was achieved by the best observer in one study (29). The results of the current meta-analysis support 3 mm as a practical size cutoff point, beneath which the sensitivity for aneurysm detection with CT angiography or MR angiography decreases sharply, from 96% to 61% and from 94% to 38%, respectively (Table 4). Furthermore, small aneurysms are seen more commonly in asymptomatic patients than in those who have an SAH: Aneurysms less than 5 mm in diameter accounted for up to a third of all the aneurysms in one study involving asymptomatic patients with nonruptured aneurysms (49); this represents a much higher proportion of small aneurysms than that in the majority of CT angiography, MR angiography, or TCD studies that have mainly included patients with aneurysms. There is relatively little information on the accuracy of noninvasive imaging in the detection of small aneurysms, but the information that is available suggests that accuracy is poor.
There is evidence that aneurysms arising from the posterior circulation (ie, intracranial vertebral arteries and basilar artery and its branches, including the posterior cerebral arteries) are at increased risk of rupture compared with aneurysms arising from the anterior circulation (ie, from the internal carotid arteries or its branches, including the posterior communicating arteries) (50,51) and that if they do rupture, the outcome is poorer (49). The detection of the subgroup of posterior circulation aneurysms is therefore particularly important.
Judging from the results of this meta-analysis, the possibility of using noninvasive modalities to screen for an aneurysm may be attractive to physicians who must deal with worried relatives of patients with an SAH. However, we believe that considerable caution is required in extrapolating data from this meta-analysis, in which the overall accuracy of CT angiography and MR angiography appeared encouraging, to the circumstances of screening, in which there is a low aneurysm prevalence—probably in the region of 5% and no more than 10%—even in at risk subgroups (51,52–54). Only one study included in this systematic review had an aneurysm prevalence in this range (40); therefore, a subgroup analysis of studies with a truly low aneurysm prevalence compared with high-prevalence studies could not be performed. Only seven of the 38 studies did not include patients known to have an intracranial aneurysm or recent acute SAH, and 20 (53%) studies focused almost exclusively on this patient group.
When there is high suspicion of an aneurysm being present, a high estimate of accuracy could result owing to observer expectation bias. The distribution of subarachnoid blood may be a strong clue to the presence and/or site of an aneurysm, as may a local hematoma or the presence of hydrocephalus. Furthermore, the results of a recent theoretic analysis suggest that increasing disease prevalence can lead to an apparent improvement in the sensitivity and specificity of a diagnostic examination, whereas previously it had been thought that increasing prevalence influenced only the predictive values—not the sensitivity and specificity—of an examination (55). In clinical practice, there usually is a single reader of a diagnostic study, whereas in some of the studies in the current systematic review, there was a consensus review by two or more readers in the analysis of accuracy, which could have resulted in a positive bias toward the noninvasive methods. These factors, coupled with publication bias and the other methodological problems outlined, suggest that the accuracy of the noninvasive examinations in the screening situation (9,10) may have been overestimated. Furthermore, it is worth noting that the NPVs were poorer than all the other results: The values per subject were 80% for CT angiography and 84% for MR angiography (Table 3). When considering the use of any examination in the context of screening, a high NPV is just as critical as high sensitivity and specificity.
In summary, there is very limited information about the accuracy of noninvasive imaging in the kind of subject likely to be screened, and it would be incorrect to assume that the accuracy will be equal to that achieved in studies with a high prevalence of aneurysms and involving symptomatic patients who have an SAH.
More data on noninvasive imaging performed in subjects with a low prevalence of aneurysms are required. When this has been done on a large scale (4,56), it has been without comparison with DSA in the majority of patients, so limited information on comparative diagnostic performance was provided. Large, prospective, blinded studies of noninvasive imaging in patients undergoing angiography but without a recent acute SAH are needed to clarify the present uncertainty, and such studies may be difficult to achieve. The small sample size in many published studies may increase the effect of random chance.
A comparison of the different methods of reviewing the images obtained with the various techniques is needed. There is some evidence that interactive workstation reconstruction and interpretation by the radiologist are more accurate than review of only film hard-copy images (57). However, workstation reconstruction is time consuming, and it is not clear whether the extra time taken is justified by the increased diagnostic yield and whether the workstation reconstruction can be adequately performed by a technician or needs to be done by the reporting radiologist.
Although technology is continually advancing, it takes time for new methods to filter through to general clinical practice. Therefore, this meta-analysis is relevant to the technologies likely to be available in most institutions now. New technologies like contrast-enhanced MR angiography with or without digital subtraction must be rigorously evaluated before they are adopted into routine clinical practice. However, as technology continues to progress, there clearly will be a role for systematic reviews in many areas of radiology, and these should be regularly updated to reflect technological advances.
In conclusion, in the populations studied in this meta-analysis, most of which had a high aneurysm prevalence, CT angiography and MR angiography had very similar diagnostic performances in the detection of intracranial aneurysms, with an accuracy of approximately 90%. The NPV per aneurysm was significantly lower than the other analysis parameters, both with CT angiography and MR angiography—67% and 77%, respectively (Table 3), and although the NPV of MR angiography was better, the 95% CI of this modality overlapped with that of the NPV of CT angiography. Sensitivity was significantly poorer for the detection of aneurysms smaller than 3 mm, and although CT angiography (61%) performed better than did MR angiography (38%), again the 95% CIs overlapped. The current data are too limited to determine confidently the accuracy of the noninvasive methods when they are used for screening. More information is needed on the optimum method for image review and on the accuracy of noninvasive imaging in subjects with a low prevalence of aneurysm.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: DSA = digital subtraction angiography, IQR = interquartile range, NPV = negative predictive value, SAH = subarachnoid hemorrhage, TCD = transcranial Doppler US
Author contributions: Guarantors of integrity of entire study, P.M.W., J.M.W.; study concepts and design, J.M.W., P.M.W.; definition of intellectual content, P.M.W., J.M.W.; literature research, P.M.W.; data acquisition, P.M.W., J.M.W.; data analysis, P.M.W., V.E.; statistical analysis, V.E.; manuscript preparation, P.M.W.; manuscript editing and review, P.M.W., J.M.W, E.V.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  L. M. Heman, L. M. Jongen, H. B. van der Worp, G. J.E. Rinkel, and J. Hendrikse Incidental Intracranial Aneurysms in Patients With Internal Carotid Artery Stenosis: A CT Angiography Study and a Metaanalysis Stroke, April 1, 2009; 40(4): 1341 - 1346. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.A.K. A. Razek and M. Castillo Imaging Lesions of the Cavernous Sinus AJNR Am. J. Neuroradiol., March 1, 2009; 30(3): 444 - 452. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P.M. White and J. Raymond Endovascular Coiling of Cerebral Aneurysms Using "Bioactive" or Coated-Coil Technologies: A Systematic Review of the Literature AJNR Am. J. Neuroradiol., February 1, 2009; 30(2): 219 - 226. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Romijn, H.A.F. G. van Andel, M.A. van Walderveen, M.E. Sprengers, J.C. van Rijn, W.J. van Rooij, H.W. Venema, C.A. Grimbergen, G.J. den Heeten, and C.B. Majoie Diagnostic Accuracy of CT Angiography with Matched Mask Bone Elimination for Detection of Intracranial Aneurysms: Comparison with Digital Subtraction Angiography and 3D Rotational Angiography AJNR Am. J. Neuroradiol., January 1, 2008; 29(1): 134 - 139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Kramer, M. J. Budoff, Z. A. Fayad, V. A. Ferrari, C. Goldman, J. R. Lesser, E. T. Martin, S. Rajagopalan, J. P. Reilly, G. P. Rodgers, et al. ACCF/AHA 2007 Clinical Competence Statement on vascular imaging with computed tomography and magnetic resonance Vascular Medicine, November 1, 2007; 12(4): 359 - 378. [PDF]
|
|
|
|
|
|
B. Lubicz, M. Levivier, O. Francois, P. Thoma, N. Sadeghi, L. Collignon, and D. Baleriaux Sixty-Four-Row Multisection CT Angiography for Detection and Evaluation of Ruptured Intracranial Aneurysms: Interobserver and Intertechnique Reproducibility AJNR Am. J. Neuroradiol., November 1, 2007; 28(10): 1949 - 1955. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Kramer, M. J. Budoff, Z. A. Fayad, V. A. Ferrari, C. Goldman, J. R. Lesser, E. T. Martin, S. Rajagopalan, J. P. Reilly, G. P. Rodgers, et al. ACCF/AHA 2007 Clinical Competence Statement on Vascular Imaging With Computed Tomography and Magnetic Resonance: A Report of the American College of Cardiology Foundation/American Heart Association/American College of Physicians Task Force on Clinical Competence and Training Developed in Collaboration With the Society of Atherosclerosis Imaging and Prevention, the Society for Cardiovascular Angiography and Interventions, the Society of Cardiovascular Computed Tomography, the Society for Cardiovascular Magnetic Resonance, and the Society for Vascular Medicine and Biology J. Am. Coll. Cardiol., September 11, 2007; 50(11): 1097 - 1114. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Papke, C. K. Kuhl, M. Fruth, C. Haupt, M. Schlunz-Hendann, D. Sauner, M. Fiebich, A. Bani, and F. Brassel Intracranial Aneurysms: Role of Multidetector CT Angiography in Diagnosis and Endovascular Therapy Planning Radiology, August 1, 2007; 244(2): 532 - 540. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Nael, J.P. Villablanca, R. Saleh, W. Pope, A. Nael, G. Laub, and J.P. Finn Contrast-Enhanced MR Angiography at 3T in the Evaluation of Intracranial Aneurysms: A Comparison with Time-of-Flight MR Angiography AJNR Am. J. Neuroradiol., November 1, 2006; 27(10): 2118 - 2121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. T. Boll, H. Bulow, K. A. Blackham, A. J. Aschoff, and B. L. Schmitz MDCT Angiography of the Spinal Vasculature and the Artery of Adamkiewicz Am. J. Roentgenol., October 1, 2006; 187(4): 1054 - 1060. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Brisman, J. K. Song, and D. W. Newell Cerebral aneurysms. N. Engl. J. Med., August 31, 2006; 355(9): 928 - 939. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Al-Shahi, P. M White, R. J Davenport, and K. W Lindsay Subarachnoid haemorrhage BMJ, July 29, 2006; 333(7561): 235 - 240. [Full Text] [PDF]
|
|
|
|
|
|
S. Sakamoto, Y. Kiura, M. Shibukawa, S. Ohba, K. Arita, and K. Kurisu Subtracted 3D CT Angiography for Evaluation of Internal Carotid Artery Aneurysms: Comparison with Conventional Digital Subtraction Angiography AJNR Am. J. Neuroradiol., June 1, 2006; 27(6): 1332 - 1337. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B.F. Tomandl, T. Hammen, E. Klotz, H. Ditt, B. Stemper, and M. Lell Bone-Subtraction CT Angiography for the Evaluation of Intracranial Aneurysms AJNR Am. J. Neuroradiol., January 1, 2006; 27(1): 55 - 59. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Hirai, Y. Korogi, H. Arimura, S. Katsuragawa, M. Kitajima, M. Yamura, Y. Yamashita, and K. Doi Intracranial Aneurysms at MR Angiography: Effect of Computer-aided Diagnosis on Radiologists' Detection Performance Radiology, November 1, 2005; 237(2): 605 - 610. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Teksam, A. McKinney, S. Casey, M. Asis, S. Kieffer, and C. L. Truwit Multi-Section CT Angiography for Detection of Cerebral Aneurysms AJNR Am. J. Neuroradiol., October 1, 2004; 25(9): 1485 - 1492. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. -R. Lu, Y. -C. Liao, J. -L. Fuh, J. -F. Lirng, and S. -J. Wang Nimodipine for treatment of primary thunderclap headache Neurology, April 27, 2004; 62(8): 1414 - 1416. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Willinek, M. Born, B. Simon, H. J. Tschampa, C. Krautmacher, J. Gieseke, H. Urbach, H. J. Textor, and H. H. Schild Time-of-Flight MR Angiography: Comparison of 3.0-T Imaging and 1.5-T Imaging--Initial Experience Radiology, December 1, 2003; 229(3): 913 - 920. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. M. Connolly, J. Huston III, R. D. Brown Jr, C. A. Warnes, N. M. Ammash, and A. J. Tajik Intracranial Aneurysms in Patients With Coarctation of the Aorta: A Prospective Magnetic Resonance Angiographic Study of 100 Patients Mayo Clin. Proc., December 1, 2003; 78(12): 1491 - 1499. [Abstract] [PDF]
|
|
|
|
|
|
D. Lardinois, W. Weder, T. F. Hany, E. M. Kamel, S. Korom, B. Seifert, G. K. von Schulthess, and H. C. Steinert Staging of Non-Small-Cell Lung Cancer with Integrated Positron-Emission Tomography and Computed Tomography N. Engl. J. Med., June 19, 2003; 348(25): 2500 - 2507. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Vogel, R. Verreault, J.-F. Turcotte, M. Kiesmann, and M. Berthel Review Article. Intracerebral Aneurysms: A Review With Special Attention to Geriatric Aspects J. Gerontol. A Biol. Sci. Med. Sci., June 1, 2003; 58(6): M520 - 524. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. K. Jayakrishnan, P. M. White, D. Aitken, P. Crane, A. D. McMahon, and E. M. Teasdale Subtraction Helical CT Angiography of Intra- and Extracranial Vessels: Technical Considerations and Preliminary Experience AJNR Am. J. Neuroradiol., March 1, 2003; 24(3): 451 - 455. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Mallouhi, S. Felber, A. Chemelli, A. Dessl, A. Auer, M. Schocke, W. R. Jaschke, and P. Waldenberger Detection and Characterization of Intracranial Aneurysms with MR Angiography: Comparison of Volume-Rendering and Maximum-Intensity-Projection Algorithms Am. J. Roentgenol., January 1, 2003; 180(1): 55 - 64. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Okahara, H. Kiyosue, M. Yamashita, H. Nagatomi, H. Hata, T. Saginoya, Y. Sagara, and H. Mori Diagnostic Accuracy of Magnetic Resonance Angiography for Cerebral Aneurysms in Correlation With 3D-Digital Subtraction Angiographic Images: A Study of 133 Aneurysms Stroke, July 1, 2002; 33(7): 1803 - 1808. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Safriel, J. Nath, J. M. Wardlaw, and P. White Noninvasive Methods in the Detection of Intracranial Aneurysms * Drs Wardlaw and White respond: Radiology, June 1, 2002; 223(3): 881 - 882. [Full Text] [PDF]
|
|
|
|
|
|
T. Abe, M. Hirohata, N. Tanaka, Y. Uchiyama, K. Kojima, K. Fujimoto, A. M. Norbash, and N. Hayabuchi Clinical Benefits of Rotational 3D Angiography in Endovascular Treatment of Ruptured Cerebral Aneurysm AJNR Am. J. Neuroradiol., April 1, 2002; 23(4): 686 - 688. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Kiyosue, M. Okahara, H. Nagatomi, T. Nakamura, S. Tanoue, and H. Mori 3.3F Catheter/Sheath System for Use in Diagnostic Neuroangiography AJNR Am. J. Neuroradiol., April 1, 2002; 23(4): 711 - 715. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Busch, F. Masuhr, K. Neubert, G. Bohner, R. Lehmann, and R. Klingebiel A Basilar Tip Aneurysm Unsuitable for Therapy: Noninvasive Visualization With Multislice Computed Tomographic Angiography Arch Neurol, March 1, 2002; 59(3): 478 - 479. [Full Text] [PDF]
|
|
|
|
 |
|
 S M Thomas The current role of catheter angiography Imaging, December 15, 2001; 13(5): 366 - 375. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P M White, E Teadsale, J M Wardlaw, and V Easton What is the most sensitive non-invasive imaging strategy for the diagnosis of intracranial aneurysms? J. Neurol. Neurosurg. Psychiatry, September 1, 2001; 71(3): 322 - 328. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. White, E. M. Teasdale, J. M. Wardlaw, and V. Easton Intracranial Aneurysms: CT Angiography and MR Angiography for Detection—Prospective Blinded Comparison in a Large Patient Cohort Radiology, June 1, 2001; 219(3): 739 - 749. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. White, J. M. Wardlaw, E. Teasdale, S. Sloss, J. Cannon, and V. Easton Power Transcranial Doppler Ultrasound in the Detection of Intracranial Aneurysms Stroke, June 1, 2001; 32(6): 1291 - 1297. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Additional articles abstracted in ACP Journal Club Evid. Based Med., May 1, 2001; 6(3): 67 - 67. [Full Text] [PDF]
|
|
|
|
|
|
W. C. Black Anatomic Extent of Disease: A Critical Variable in Reports of Diagnostic Accuracy Radiology, November 1, 2000; 217(2): 319 - 320. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
