| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Evidence-based Practice |
1 From the Division of Clinical Neurosciences, University of Edinburgh, Western General Hospital, Crewe Rd, Edinburgh, EH4 2XU, Scotland (J.M.W.); and Department of Neurology, Klinikum Mannheim, University of Heidelberg, Mannheim, Germany (O.M.). Received February 12, 2004; revision requested April 15; revision received May 31; accepted July 1. Supported by Chest, Heart and Stroke Scotland grant RES000/7. O.M. supported by the Department of Neurology, Klinikum Mannheim, University of Heidelberg, Mannheim, Germany. Address correspondence to J.M.W. (e-mail: joannawardlaw@ed.ac.uk).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: A systematic review of the literature was conducted by using Cochrane Stroke Group methodology to identify studies published between 1990 and 2003 that were performed to assess interobserver agreement about early signs of infarction on CT scans obtained within 6 hours after onset of stroke symptoms and determine the relation of early signs of infarction to clinical outcome, including any interactive effect of thrombolysis. Interobserver agreement was measured with the 
statistic, sensitivity, and specificity. The relation of early signs to clinical outcome with or without thrombolysis was assessed with calculated odds ratios and 95% confidence intervals.
RESULTS: In 15 studies of interobserver agreement (median of 30 CT scans and six raters), the prevalence of all early infarction signs was 61% ± 21 (standard deviation). Interobserver agreement ( statistics) ranged from 0.14 to 0.78 for any early infarction sign. The mean sensitivity and specificity for detection of early infarction signs with CT were 66% (range, 20%–87%) and 87% (range, 56%–100%), respectively. Experience improved detection, but knowledge of symptoms did not. In 15 studies of early infarction signs and outcome (including seven thrombolysis trials) in 3468 patients, any early infarction sign increased the risk of poor outcome (odds ratio, 3.11; 95% confidence interval: 2.77, 3.49). Two studies that sought interaction between early infarction signs and thrombolysis found no evidence that thrombolysis given in the presence of early infarction signs resulted in worse outcome than that due to early signs alone.
CONCLUSION: Further work is required to determine which signs are most reliably detected, whether scoring systems help to improve detection, and whether any early infarction sign should influence decisions concerning thrombolysis.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The signs evaluated so far are mainly located in the territory of the middle cerebral artery and are defined in semiquantitative terms (eg, as thirds of the middle cerebral artery) (12,22), in qualitative terms (eg, obscuration of lentiform nucleus, cerebral sulcal effacement), or in terms of a pathologic vascular condition (eg, hyperattenuation of the middle cerebral artery) (23). These early signs, however, are acknowledged to be difficult to identify and were not used as exclusion criteria in the National Institute of Neurological Disorders and Stroke (NINDS) trial (24), which is still the only trial that showed better outcomes in patients treated with thrombolysis after stroke.
Thrombolysis is not widely implemented for treatment of acute ischemic stroke (25,26), and uncertainty about the interpretation of early infarction signs on CT scans may be slowing its implementation. Magnetic resonance (MR) diffusion-weighted imaging and perfusion imaging may help to identify patients for thrombolysis, but there is still a need to determine how best to use CT, as CT is likely to remain the primary modality for assessment of stroke because of its wide availability, ease of use, and speed.
Thus, the purpose of our study was to review systematically all reported early CT signs in acute ischemic stroke to determine interobserver agreement and the relationship between early CT signs and patient outcome with or without thrombolysis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
|
|
statistics) were readily available or in which the reader sensitivity and/or agreement rate was measured and compared with a defined reference standard. Studies in which investigators compared CT signs and outcome were restricted to those that included 10 or more patients in whom clinical outcome was assessed at least 1 month after stroke and as the functional outcome (death or dependency) according to a recognized standard scale (eg, Rankin or Barthel) and for whom it was stated whether thrombolysis had been used or not. Because findings in studies with small sample sizes are less statistically robust or generalizable, studies with fewer than 10 subjects were excluded. We also excluded studies published before 1990, because early signs of infarction were not well recognized and thrombolysis was not widely used at that time and because CT technology has since changed. Studies of stroke in children and studies in which only radiologic markers of outcome (eg, infarct volume) were given, without clinical outcome data, likewise were not included.
Identified studies that met the initial criteria were then formally assessed for eligibility by using a standardized assessment form. We did not implement a quality score as an inclusion or exclusion criterion for the present review because of the great heterogeneity of the studies, but we did exclude studies in which information mandatory for the analysis was lacking.
Data Extraction
Both authors independently extracted data by using a standardized assessment form (Table 2) to record the number of observers, their years of experience in reading CT scans, the number of scans and the method of review (film image or electronic image viewed at a picture archiving and communication system workstation), demographic details about the patients (age, type of stroke, severity of stroke), the early infarction signs sought and the definitions of those signs, and statistical measures of interobserver agreement ( statistic, sensitivity, specificity, or accuracy). Observers were classified as expert (neuroradiologists with a major interest in stroke, who provided the reference reading of CT scans in trials such as the European Cooperative Acute Stroke Study trial [ECASS I]), experienced (radiologists or physicians at the consultant or board-certified level who could reasonably be expected to have more than 5 years of experience in reading stroke CT scans in clinical practice), or novice or less experienced (trainees, neurologists, physicians developing a specialization in stroke, or radiologists learning to interpret stroke CT scans but with less than 5 years of experience).
|
Both authors independently checked all of the extracted data against the publications twice, to ensure correct and complete data extraction. Any discrepancies in extracted data were discussed, and disagreements were resolved with consensus.
Statistical Analysis
We calculated the following descriptive statistics: the mean or median number of scans and observers per study, the total number of scans per observer, and the proportion of scans with early infarction signs. Data about interobserver agreement ( statistic, sensitivity, specificity, or accuracy) were extracted from the published data or, when possible, calculated from the data given. We expressed statistics, sensitivity, and specificity as the range of values extracted from the studies, and we tabulated these data by using spreadsheet software (Excel 2000, version 9.0.6926 SP-3; Microsoft, Seattle, Wash). We also merged the early signs described in the primary publications into three broader categories (those describing the lesion extent, those describing brain attenuation change, and those describing mass effect) and, finally, into two primary categories (any versus no early signs). Further, we summarized , sensitivity, and specificity values as the range and mean by using the same spreadsheet software. We planned to examine the effect of observer experience, of knowledge of symptoms, and of the prevalence of early infarction signs on interobserver agreement if enough data were available.
In the studies of early infarction signs and outcome without or with thrombolysis, we determined the proportion of patients with early infarction signs and good or poor outcomes and calculated the odds ratio for a poor outcome in the presence of each particular sign. We calculated summary odds ratios by using software (Review Manager, version 4.2 for Windows 2000; The Cochrane Collaboration, Oxford, England). In randomized trials of thrombolysis versus control, we planned to examine whether the effect of thrombolytic therapy on clinical outcome was different in patients in whom early infarction signs were present, compared with those in whom such signs were absent.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
In the 15 studies included in our analysis (Table 3), 1300 scans were analyzed by 709 readers. One study with 20 scans (532 readers) took place during the training of ECASS II trial investigators (36). In another study, the difference in performance between readings by local investigators and those by three expert observers in ECASS II was assessed (20). The mean number of CT scans interpreted per study was 87 (median, 30), and that of readers was 51 (median, six). After the exclusion of the data from the ECASS II training study, the mean number of CT scans per study among the remaining 14 studies was 91 (median, 30), with a mean of 14 readers per study (median, five).
|
Printed hard-copy CT scans were used in all studies except one in which scans were viewed at a picture archiving and communication workstation (31). Between one and five early CT signs were analyzed, including hypoattenuation, swelling, hyperattenuation of the middle cerebral artery, loss of definition between gray matter and white matter, loss of basal ganglia outline, loss of the insular ribbon, infarction of more or less than one-third of the middle cerebral artery territory, and cortical sulcal effacement. In some studies, hemorrhage, previous infarction, and/or signs that may mimic cerebral infarction also were considered (30,33,38). Explicit definitions of early signs of infarction were given in only eight of the 15 studies.
An early sign of infarction or vascular pathologic condition was present on a mean of 52 (61% ± 21) of 85 scans. The reference standard in most cases was reading by a neuroradiologist with follow-up imaging (virtually always with CT). Investigators in one study used angiography-proved vessel occlusion to confirm a CT finding of hyperattenuation of the middle cerebral artery (HMCAO sign) (29), and in three other studies, diffusion-weighted MR images were available for comparison with CT scans (27,28,35).
In nine studies, readers were blinded to clinical features; in three other studies, they were not. In another three studies, readers were first blinded and then unblinded (34,37,38) to clinical features. Four studies gave information on observer reliability according to prior experience of the observers (28,33,36,38).
Interobserver agreement was expressed in some studies as a value and in others as a percentage of agreement between an expert reader and a less experienced reader or as sensitivity and specificity. These variations made comparisons between studies very difficult. For interobserver agreement expressed as (agreement beyond chance), values were wide ranging; for example, the range of values was 0.14–0.78 for the detection of any early sign of infarction (24,27,28,43,44,46), with the lowest and highest values in that range indicating slight and substantial agreement, respectively, according to Landis and Koch (49). The highest levels of agreement were those for the HMCAO sign, for which values ranged from 0.36 to 1.00. In general, infarction signs had higher specificity than sensitivity. The mean sensitivity for infarction sign detection was 66% (range, 20%–87%), and the mean specificity was 87% (range, 56%–100%) (Table 3).
Investigators in six studies examined individual components of early infarction signs in terms of sensitivity and specificity (37,38) and statistics (24,37,43,46). In general, in studies in which parenchymal signs were evaluated along with the hyperattenuated middle cerebral artery sign, interobserver agreement was worst for hypoattenuation and best for the hyperattenuated artery sign, with two exceptions (32,43). The values for hypoattenuation ranged from 0.30 to 0.53; those for mass effect, from 0.41 to 0.58; those for loss of differentiation between gray matter and white matter, from 0.20 to 0.35; those for the one-third middle cerebral artery sign (37,43,50), from 0.37 to 0.62 (sensitivity, 44%–87%); and those for the hyperattenuated artery sign, from 0.36 to 1.00.
There were too few data or details given to allow us to reliably determine the effects of observer experience, of knowledge of the patient’s symptoms, and of the proportion of scans with early signs on interobserver agreement. In one study, it was suggested that experienced readers performed more consistently than less experienced readers and rarely saw signs that were not present, whereas less experienced readers often saw signs that were not present or were false signs of infarction (38). In another study, experienced readers generally performed better than novices in detecting any, as well as specific, infarction signs, but there were some inconsistencies (28). In one study, experienced readers were less likely to misinterpret hemorrhage as calcification and were slightly better at detecting early infarction signs (33). The ECASS trial investigators found that training resulted in a significant increase in the detection of any early infarction signs (P < .0001) (36). Compared with the local trial investigators, the expert CT readers in ECASS II had significantly higher sensitivity (P < .001) and accuracy (P < .001) but not specificity (P > .05) for detection of hypoattenuation (20).
In the three studies in which observers were tested while blinded or unblinded to the side of the brain affected by infarction, knowledge of the side of infarction or of symptoms did not change interobserver agreement significantly (P > .05) (34,37,38).
Studies of Early CT Signs and Clinical Outcome
Fifteen studies met our criteria. Five were secondary analyses of results from seven previously published randomized controlled trials of thrombolysis (NINDS trial, ECASS I and II, Australian Streptokinase [ASK] trial, Multicenter Acute Stroke Trial [MAST]–Italy, MAST–Europe, and Prourokinase in Acute Stroke Trial) (Table 4). The total number of patients in these 15 studies was 3481. The results from some trials were analyzed more than once with regard to different CT signs and outcome parameters (8,10,19,51); however, when a subsequent publication reported the results of a repeat analysis of CT infarction sign data, we did not include data from the primary publication, to avoid duplicate counting. The remaining 10 studies were observational studies in 947 patients, 680 (72%) of whom underwent either local or systemic thrombolysis. Four further studies that initially appeared to contain information about early infarction signs were excluded. One study (46) was excluded because it was not possible to separate the data from patients who underwent CT less than 6 hours after infarction from the data from patients who underwent CT between 6 and 14 hours after infarction. A second study (13) was excluded because the outcomes were not reported according to the sign of infarction; a third (17) was excluded because all patients with early infarction signs had been excluded from the initial analysis; and a fourth (7) was excluded because clinical follow-up was performed only 7 days after infarction.
|
Baseline patient data, when mentioned, were scant and mainly included age, sex distribution, and baseline stroke severity. In nine studies, no baseline patient data were provided.
Observers in six studies analyzed CT scans that were obtained within 3 hours of stroke onset. One study included CT scans obtained up to 7 hours after stroke (16), and most of the others included scans obtained up to 6 hours after stroke. Early sign of infarction was present on a mean of 57% ± 22 (range, 17%–100%) of CT scans. Early infarction signs were mostly defined qualitatively (ie, descriptively; see Table 1 for details). Only in one study did investigators measure tissue attenuation change directly at the CT console (4).
Early infarction signs and functional outcome.—Early infarction signs were correlated with clinical outcome after a defined follow-up period of at least 1 month in 14 studies. An attempt was made to calculate an odds ratio for death or dependency at long-term follow-up (a) if any early sign was present or (b) if the ischemic damage extended through more or less than one-third of the corresponding middle cerebral artery territory or (c) if the artery was hyperattenuated or not (Fig 2). Early infarction signs increased the risk of poor functional outcome (overall odds ratio, 3.11; 95% confidence interval [CI]: 2.77, 3.49), but these analyses did not determine any interaction with thrombolysis.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
The sensitivity and specificity of early infarction signs on CT scans are generally not good, and the values for interobserver agreement, when given, usually reflect only poor to moderate agreement. It would appear, however, that the sensitivity of early infarction signs on CT scans is poorer than the specificity; in other words, signs are not generally well detected, but when seen, they are likely to be present. Hyperattenuation of the middle cerebral artery was in general more reliably detected than were changes in brain parenchymal attenuation or swelling, but there were insufficient data to determine whether any of the parenchymal indicators of early infarction were better detected than others. Little information was available about the effects of knowledge of the patient’s symptoms, use of scoring systems, or experience or inexperience (or training or lack of training) on reader performance, but the available information suggests that the reader’s knowledge of symptoms did not affect his or her detection of early infarction signs, that experienced readers were more consistent and accurate than inexperienced or less experienced readers, and that training improved performance. Almost all of the data come from studies in which hard copies of CT scans were read; yet, in routine clinical practice, scans increasingly are viewed only at picture archiving and communication workstations. In only one study was scan interpretation evaluated at a workstation (29), and in none were both hard-copy and workstation image evaluations assessed; consequently, there are no available data about any effect of the viewing method on interobserver agreement.
For the relationship between early infarction signs and outcome and the interaction with thrombolysis, data were likewise scant. There were multiple publications of results from the same studies, and we may not have been able to completely avoid duplicate counting. There was a consistent pattern indicating that early infarction signs were associated with poorer clinical outcomes. However, in the only two studies in which such an association was sought, no evidence was found of any interaction between early infarction signs, clinical outcome, and thrombolysis with either streptokinase given up to 6 hours or alteplase given up to 3 hours after stroke (3,51). In other words, there was no evidence that, in the presence of early infarction signs, thrombolysis worsened functional outcome.
Does this mean that patients with early infarction signs can safely be treated with thrombolysis? There are several problems in interpreting these data. The trials were mostly conducted prior to 1995, and both CT technology and the awareness of early infarction signs have improved since then. Findings of hypoattenuation in 1990–1995 may well have involved a much less subtle change in attenuation than did those in 2004. Changing awareness and improving technology may have increased the detection rate, but unfortunately it was not possible to determine this from the data available. The analyses were based predominantly on centralized readings of scans by expert observers, not on what the investigators at the participating study hospital thought, and less experienced readers may fail to see some early signs. Patients with very obvious early signs may not have been included in the trials; although early CT infarction signs were not specified as exclusion criteria in any of the trials that contributed to these analyses (NINDS and the streptokinase trials), the investigators at the participating hospitals may have excluded any patient who they thought showed early infarction signs, because of a prior belief that early signs increase the hazard of thrombolysis. Without a standardized reinterpretation of the existing CT scans by a larger group of expert readers to determine the prevalence and severity of signs in each trial cohort, it is difficult to make further comment.
Our systematic review has highlighted the general lack of definitions for early infarction signs. Investigators in a substantial proportion of the studies did not define the infarction signs they sought. Indeed, one of the most widely used signs, the "less than or more than one-third of the middle cerebral artery" sign, seems to have been clearly defined only in 2001, a decade after the sign came into use and after most of the studies in this review had been published (12). Investigators in future trials clearly need to be much more specific about their definition of early signs and interpretation of CT scans to improve observer reliability and aid interpretation of any interactions of thrombolysis treatment.
Several further studies that were published after the end of our search of the literature are relevant to this discussion. Hill et al (54) described a further evaluation of their ASPECTS system, which was applied in the PROACT II study of CT scans. Patients with an ASPECTS value (1) of more than 7 were three times more likely to be independent after administration of recombinant prourokinase than were control patients (risk ratio, 3.2; 95% CI: 1.2, 9.1). Patients with an ASPECTS value of 7 or less were no more likely to be independent than were control patients (risk ratio, 1.0; 95% CI: 0.6, 1.9). The investigators did not specifically examine interactions among treatment, early infarction score, and outcome. Mak et al (55) compared the sign of hypoattenuation in one-third of the middle cerebral artery with the ASPECTS value determined in 80 consecutive patients admitted to the hospital for evaluation with CT within 6 hours after infarction. There were more patients with an ASPECTS value of 7 or less than there were with infarction signs involving more than one-third of the middle cerebral artery territory. The percentage of agreement between five blinded observers was 71% for involvement of more than one-third of the middle cerebral artery territory ( = 0.49) and 42% for an ASPECTS value of 7 or less ( = 0.34). Unblinding did not alter observer agreement. After various adjustments for prevalence and bias, Mak et al concluded that the "one-third of the middle cerebral artery" method was more reliable for detection of important early infarction signs at CT within 6 hours after stroke (ie, was more specific), whereas the ASPECTS value enabled detection of important early infarction signs in a higher proportion of the early scans (ie, was more sensitive). This highlights a further problem with many of the data included in our review: It is difficult to ascertain the generalizability of the interobserver agreement results, as many publications did not include information about the patient population in which the CT scans were obtained. Thus, some signs may have appeared to perform better (or worse) than they would in actual clinical practice, because they were present on a high (or low) proportion of the scans included in the particular study. Future studies of this subject should specify the population in which CT scans were obtained.
What are the limitations of our review? It was difficult to be certain that we had obtained all the relevant literature, as search methods are less well developed for diagnostic tests than for studies of treatments. We may therefore have inadvertently omitted studies by failing to find them. We do not know how many studies of early infarction signs remain unpublished, or what effect publication bias might have had on the results of our review. It was also difficult to be certain that we had not inadvertently included some data twice. It was sometimes very difficult to be certain that two publications by a similar group of authors actually contained data from different patients. Most of the included studies were conducted with somewhat artificial conditions that may have affected the results. For example, interobserver agreement among readers presented with a batch of scans known to be from acute stroke patients may be better than that in actual day-to-day practice, in which an acute stroke patient may undergo CT on the same day as do patients with other pathologic conditions. In most studies of early infarction signs and outcome, the scans were read by experienced observers. The interobserver agreement data from the first part of our review suggest that many of these signs would have been overlooked if less experienced observers had read the scans.
In summary, further work is necessary to improve the definition and standardization of early infarction signs at CT and to determine which signs are most reliably detected, whether scoring systems help and should be more widely used, whether early infarction signs should be taken into account when decisions are made concerning thrombolytic treatment soon after stroke, and, if so, what degree of early infarction sign is important. Large-scale observer reliability studies that include readers with a wide range of degrees of expertise and backgrounds are needed to evaluate not just the interpretation of expert observers but also that of local physicians and radiologists who are likely to be making decisions about stroke treatment. Reevaluation of existing thrombolysis trial data may further improve understanding of the relationship between early infarction signs and outcome with thrombolysis, but further data collected in modern settings in large trials also are necessary to provide robust answers about these complex interactions.
|
FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantor of integrity of entire study, J.M.W.; study concepts and design, J.M.W.; literature research, J.M.W., O.M.; data acquisition and analysis/interpretation, J.M.W., O.M.; statistical analysis, J.M.W., O.M.; manuscript preparation, definition of intellectual content, editing, and revision/review, J.M.W., O.M.; manuscript final version approval, J.M.W.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. M. Wardlaw Surrogate Outcomes: A Cautionary Note Stroke, April 1, 2009; 40(4): 1029 - 1031. [Full Text] [PDF]
|
|
|
|
 |
|
 T Kharitonova, M Thoren, N Ahmed, J M Wardlaw, R von Kummer, L Thomassen, N Wahlgren, and for the SITS investigators Disappearing hyperdense middle cerebral artery sign in ischaemic stroke patients treated with intravenous thrombolysis: clinical course and prognostic significance J. Neurol. Neurosurg. Psychiatry, March 1, 2009; 80(3): 273 - 278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kobayashi, J. M. Wardlaw, R. I. Lindley, S. C. Lewis, P. A.G. Sandercock, A. Czlonkowska, and on behalf of the IST-3 Collaborative Group Oxfordshire Community Stroke Project Clinical Stroke Syndrome and Appearances of Tissue and Vascular Lesions on Pretreatment CT in Hyperacute Ischemic Stroke Among the First 510 Patients in the Third International Stroke Trial (IST-3) Stroke, March 1, 2009; 40(3): 743 - 748. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. M. de Lucas, E. Sanchez, A. Gutierrez, A. G. Mandly, E. Ruiz, A. F. Florez, J. Izquierdo, J. Arnaiz, T. Piedra, N. Valle, et al. CT Protocol for Acute Stroke: Tips and Tricks for General Radiologists RadioGraphics, October 1, 2008; 28(6): 1673 - 1687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Pettersen, Y. Haig, P. H. Nakstad, and T. B. Wyller Subtypes of urinary incontinence after stroke: relation to size and location of cerebrovascular damage Age Ageing, May 1, 2008; 37(3): 324 - 327. [Full Text] [PDF]
|
|
|
|
|
|
A. Ogawa, E. Mori, K. Minematsu, W. Taki, A. Takahashi, S. Nemoto, S. Miyamoto, M. Sasaki, T. Inoue, and for The MELT Japan Study Group Randomized Trial of Intraarterial Infusion of Urokinase Within 6 Hours of Middle Cerebral Artery Stroke: The Middle Cerebral Artery Embolism Local Fibrinolytic Intervention Trial (MELT) Japan Stroke, October 1, 2007; 38(10): 2633 - 2639. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. P. Adams Jr, G. del Zoppo, M. J. Alberts, D. L. Bhatt, L. Brass, A. Furlan, R. L. Grubb, R. T. Higashida, E. C. Jauch, C. Kidwell, et al. Guidelines for the Early Management of Adults With Ischemic Stroke: A Guideline From the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups: The American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists. Circulation, May 22, 2007; 115(20): e478 - e534. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. P. Adams Jr, G. del Zoppo, M. J. Alberts, D. L. Bhatt, L. Brass, A. Furlan, R. L. Grubb, R. T. Higashida, E. C. Jauch, C. Kidwell, et al. Guidelines for the Early Management of Adults With Ischemic Stroke: A Guideline From the American Heart Association/ American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups: The American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists Stroke, May 1, 2007; 38(5): 1655 - 1711. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Wardlaw, A. J. Farrall, D. Perry, R. von Kummer, O. Mielke, T. Moulin, A. Ciccone, M. Hill, and for the Acute Cerebral CT Evaluation of Stroke Stu Factors Influencing the Detection of Early CT Signs of Cerebral Ischemia: An Internet-Based, International Multiobserver Study Stroke, April 1, 2007; 38(4): 1250 - 1256. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Dzialowski, M. D. Hill, S. B. Coutts, A. M. Demchuk, D. M. Kent, O. Wunderlich, and R. von Kummer Extent of Early Ischemic Changes on Computed Tomography (CT) Before Thrombolysis: Prognostic Value of the Alberta Stroke Program Early CT Score in ECASS II Stroke, April 1, 2006; 37(4): 973 - 978. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.O. McCarron, C. Sands, and P. McCarron Quality assurance of neuroradiology in a District General Hospital QJM, March 1, 2006; 99(3): 171 - 175. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Tanaka, T. Ueguchi, E. Shimosegawa, N. Sasaki, T. Johkoh, H. Nakamura, and J. Hatazawa Effect of CT Acquisition Parameters in the Detection of Subtle Hypoattenuation in Acute Cerebral Infarction: A Phantom Study AJNR Am. J. Neuroradiol., January 1, 2006; 27(1): 40 - 45. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Hirano, T. Yonehara, Y. Inatomi, Y. Hashimoto, and M. Uchino Presence of Early Ischemic Changes on Computed Tomography Depends on Severity and the Duration of Hypoperfusion: A Single Photon Emission-Computed Tomographic Study Stroke, December 1, 2005; 36(12): 2601 - 2608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Allard, P. R. Burkhard, P. Lescuyer, J. A. Burgess, N. Walter, D. F. Hochstrasser, and J.-C. Sanchez PARK7 and Nucleoside Diphosphate Kinase A as Plasma Markers for the Early Diagnosis of Stroke Clin. Chem., November 1, 2005; 51(11): 2043 - 2051. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
