Noncontrast conventional computed tomography in the evaluation of acute stroke

Noneontrast Conventional Computed Tomography in the Evaluation of Acute Stroke

Sean P. Symons, Sean P. Cullen, Ferdinando Buonanno, R. Gilberto Gonz~lez, and Michael H. Lev

C ONVENTIONAL computed tomography (CT) scanning is currently the first-fine diagnostic

test for the emergency evaluation of acute stroke patients because of its speed, widespread availability, low cost, and accuracy in detecting extra-axial and parenchymal hemorrhage. 1-6 CT scans can detect ischemic brain regions within 6 hours of stroke onset (hyperacute). Indeed, the sensitivity of such early detection likely exceeds that of conventional T2- weighted magnetic resonance imaging. 7-12 The identification of iscbemic brain tissue by a CT scan not only defines regions likely to infarct but may predict outcome and response to intravenous (IV) or intra- arterial (IA) thrombolyfic therapy.13 Most commonly, a CT scan is used to exclude hemorrhage and establish infarction size.

This article reviews the role of conventional noncontrast CT scanning in the evaluation of hyperacute stroke patients, including discussion of data acquisition and image interpretation techniques, the physical basis of CT imaging findings, and the accuracy and clinical usefulness of CT scanning.

TECHNIQUE

Noncontrast CT scanning of acute stroke patients presenting to the Massachesetts General Hospital, Boston, MA, is typically performed in the emergency department using a LightSpeed Multislice Helical CT scanner (GE Medical Sys- tems, Milwaukee, WI). The following helical scan technique is routinely used: 140 kV, 170 mA, 5-mm slice thickness, 5-ram image spacing, 3:1 pitch, 0.8-second rotation time, and 22-cm field of view. Coverage is from the skull base to the vertex, using contiguous axial slices. The scans are obtained in a head holder to optimize registration between the noncontrast and contrast images for patients that undergo CT angiography immediately after conventional CT. The previously mentioned scanning parameters were chosen as a tradeoff between maximizing resolution and image contrast of the brain parenchymal images (most notably in the posterior fossa) and minimizing both scan time and patient radiation dose. To this end, we discour- age the use of high-speed helical technique (pitch of 6 or more) for brain imaging because the degree of helical reconstruction artifact through the pos-

terior fossa may produce images of unacceptable diagnostic quality. One hundred seventy milliam- peres is specifically chosen to take advantage of the small radiograph tube focal spot setting. It is noteworthy that when using this scan protocol, the resulting axial 5-ram thick source images can be reformatted to 2.5-ram thick slices in either bone or soft-tissue algorithm should the need arise.

PHYSICAL BASIS OF IMAGING FINDINGS

The pathophysiology underlying the changes detected by CT when viewing hyperacute stroke patients is outlined in a review by Marks. 4 Atten- uation values on CT scan images, measured in Hounsfield units (HU), reflect tissue density. Low- density substances, such as air, fat, and water, appear dark, whereas high-density substances, such as bone, contrast, or hemorrhage, appear bright. The finding of early parenchymal hypodensity (minutes to hours), secondary to cytotoxic edema, is caused by lactic acidosis and failure of cell membrane ion pumps because of inadequate adenosine triphosphate supply. TM Hypodensity secondary to vasogenic edema, which is caused by loss of tight junction endothelial cell integrity, appears later (hours to days) and only if residual or resumed perfusion is present, x4 The HU attenuation of acutely ischemic brain parenchyma is di- rectly proportional to the degree of edema. For every 1% increase in tissue water content, radiograph attenuation decreases by 3% to 5%, which corresponds to a drop of approximately 2.5 HU on CT imaging. 4,15,16

OPTIMAL IMAGE REVIEW

The window and level settings (measured in HU) used for CT image review are potentially important variables in stroke detection because the

From the Neuroradiology Division and Stroke Service, Mas- sachusetts General Hospital and Harvard Medical School, Boston, MA.

Address reprint requests to R. Gilberto Gonzdlez, MD, PhD, Neuroradiology, GRB 285, Fruit Street, Massachusetts General Hospital Boston, MA 02114-2696.

Copyright 2002, Elsevier Science (USA). All rights reserved. 0037-19832/02/3703-0004535.00/0 dok l O. 1053/sroe.2002.34559

Seminars in Roentgenology, Vo[ 37, No 3 (July), 2002: pp 185-191 185

186 SYMONS ET AL

Fig 1. Identical noncontrast CT scans of a patient presenting with acute left NICA infarct symptoms with standard (1), and narrow window and level settings (2). Note the increased conspicuity of the left lentiform nucleus infarct with the narrow window and level settings. The infarct was confirmed by diffusion-weighted imaging.

decreases in CT attenuation accompanying early stroke are small. These settings influence lesion conspicuity and diagnostic accuracy in diseases other than stroke. For example, in the CT angiographic evaluation of severe carotid artery stenosis, optimal window and level viewing parameters are required for precise luminal diameter measurement. 17-19 In abdominal CT imaging, the routine use of narrow liver windows has been recom- mended to improve soft-tissue contrast, thereby increasing the conspicuity and detection of subtle liver lesions. 2° It has been shown that the CT detection of acutely ischemic brain parenchyma is also facilitated by soft-copy image review using narrow, nonstandard window and level settings. 2~ The settings are modified to accentuate the small attenuation differences between normal and ischemic gray and white matter (Fig 1). 2~ Using nonstandard, narrow window and level review settings in the CT detection of hyperacute stroke has been shown to increase the sensitivity of stroke detection from approximately 57% (using standard review parameters) to over 70%. 2~ The use of interactive, soft-copy window and level settings does not significantly increase the time required for scan interpretation over that required using the standard, preset settings. This method of image review has not previously been emphasized, possibly in part because of the historical unavailability of a conve- nient, digital, soft-copy image review capability. Given the marked increased use of picture archiving and communication system workstations, however, there is the potential for increasing diagnostic

accuracy through interactive image interpretation without increasing the time or expense required to film and review hard-copy images at multiple window and level settings. 22-24

The idea of exploiting the subtle difference in density between normal and acutely edematous brain tissue to more sensitively detect stroke is not new. A 1997 report described an automated postprocessing method for the CT detection of middle cerebral artery (MCA) territory infarcts using his- tographic analysis of density values. 25 Attenuation values for each CT slice were recorded for each hemisphere, and a density-difference diagram was calculated by digital subtraction of the histogram of the left hemisphere from that of the right. Using density-difference analysis (DDA), the detection rate for infarcts increased from 61% to 96%. However, unlike the use of narrow soft-copy window and level review settings, DDA requires specialized postprocessing equipment and additional interpretation time.

More recently, Maldjian et a126 have exploited the idea that acutely ischemic regions have subtle hypoattenuation to develop a semiautomated segmentation program that can be used to aid in detecting and delineating subtle ischemic regions.

ACCURACY AND PROGNOSTIC VALUE

In general, the presence of an ischemic hypodensity in a patient with signs and symptoms of hyperacute stroke signifies irreversible infarction. This is consistent with our clinical experience and has been confirmed in a number of recent studies.27,28

Sensitivity values for the CT detection of acute ischemic stroke vary in the literature. These differences are largely caused by the variance in study design and focus, the vascular territory in question, and the generation of CT scanner used. Because control groups without stroke are not commonly included as subjects in published reports, the spec- ificity of CT scanning in stroke detection is not well establishedY Early generation CT scanners often failed to detect stroke until 2 days after ictus? 9 In contrast to this, a 1989 study of patients with MCA infarction reported that 70% of CT examinations obtained within 4 hours of stroke onset showed focal decreased attenuation consistent with tissue ischemia. 3°

CT findings that are useful in the detection of acute ischemia include (1) focal parenchymal hy-

CONVENTIONAL COMPUTED TOMOGRAPHY 187

Fig 4. Noncontrast CT scan of a patient presenting with acute left MCA infarct symptoms showing cortical swelling with sulcal effacement and loss of gray-white differentiation. The infarct was confirmed by diffusion-weighted imaging. A follow-up noncontrast CT scan shows the evolving left MCA infarct.

Fig 2. Four contiguous axial noncontrast CT scans of a patient presenting with acute right MCA infarct symptoms showing hypodensity of the insular ribbon. The infarct was confirmed by diffusion-weighted imaging,

podensity, most notably of the insular ribbon (Fig 2) or lenticular nuclei (Fig 3) for MCA territory strokes; (2) cortical swelling with sulcal efface- merit (Fig 4); (3) loss of gray, white matter differentiation (Fig 4); and, less reliably, (4) the hyperdense MCA sign (Fig 5) or MCA dot sign (Fig 6). 4'7'8'13'31-33

Of these, parenchymal hypodensity is the most important. Tomura et al 7 reported a 92% rate of lentiform nucleus obscuration within 6 hours of MCA stroke. In another series of patients studied within 5 hours of angiographically proven MCA occlusion before intravenous (IV) thrombolysis, early CT showed hypodensity in 81%, brain swelling in 38%, and a hyperdense MCA in 47%. 13

Cortical swelling and sulcal effacement have not been as useful indicators of acute infarction; 1 series found no swelling in the first 3 hours after ictus, and another found swelling in only 23% of patients imaged within 3 hours. 7,8

Hyperdense middle cerebral artery sign (HMCAS) refers to increased density of an MCA segment compared with the contralateral normal side caused by the presence of an acute thrombus. HMCAS, first described by Pressman and Tourje 31 in 1987, is a specific but not sensitive indicator of acute MCA stroke; estimates of its incidence range from 1% to 50%. 4 Hyperdense attenuation of other arteries, such as the basilar, may also indicate the presence of thrombus (Fig 7). It has been reported that an M1 MCA HU absolute density of greater than 43 and a HU ratio greater than 1.2 between the

Fig 3. Noncontrast CT scans of 2 patients presenting with acute NICA infarct symptoms showing hypodensity of the lenticular nuclei. The infarcts were confirmed by diffusion- weighted imaging,

Fig 5. Noncontrast CT scans of 6 different patients presenting with acute MCA infarct symptoms showing hyperdense middle cerebral artery signs indicating acute M1 thrombus. These were angiographically confirmed,

188 SYMONS ET AL

Fig 6. Noncontrast CT scans of 2 different patients presenting with acute MCA infarct symptoms showing MCA dot signs indicating acute M2 thrombus. These were angiographically confirmed.

affected and unaffected MCAs is predictive of a true HMCAS. 32 Hyperdensity of the sylvian MCA branches has been described as an MCA dot sign and is indicative of M2 or M3 thrombus. 33 The MCA dot sign can be seen with or without a HMCAS. The MCA dot sign has been described as being approximately 52% sensitive and 92% specific of acute embolic thrombus. 34

These early CT changes of ischemia not only facilitate stroke detection but can also help predict prognosis and response to thrombolytic therapy. 10'13'35-39 In a study that revealed HMCAS in 18 of 55 (33%) patients presenting within 90 minutes of stroke onset, HMCAS predicted poor outcome after IV thrombolysis; a National Institute of Health stroke scale score of greater than 10 was an even better predictor of outcome. 4° Another study reported that early low attenuation in greater than 50% of the MCA territory is predictive of up to 85% mortality, with a poor outcome in survi- vors. 13

In the first European Cooperative Acute Stroke Study (ECASS I), a double blind placebo-con- trolled trial of IV thrombolysis administered within 6 hours of stroke onset, patients with initial CT findings of a greater than one-third MCA territory hypodensity or sulcal effacement were shown to have had an increased risk of fatal parenchymal hemorrhage after treatment. ~,41 Because of this, these findings are considered by some to be con- traindications to thrombolytic therapy. 1°,36,37,4~,42 The ECASS II study confirmed the importance of the extent of hypodensity as a major risk factor for severe hemorrhagic transformation? 3

In the National Institute of Neurological Disor- ders and Stroke (NINDS) study, which showed a

benefit of IV thrombolysis administered within 3 hours of stroke onset, there was a trend toward improved outcome despite the presence of early CT hypodensityS In this trial, patients were randomized to IV recombinant tissue plasminogen activator (rt-PA) versus placebo. Patients with hemorrhage on initial CT were excluded from the NINDS study, a less strict entry criterion than was applied for the ECASS I trial, in which patients with either early hemorrhage or significant parenchymal hypodensity on initial CT were excluded. In neither study was angiographic proof of a vascular occlusion amenable to thrombolytic treatment obtained before the administration of rt-PA.

It is controversial that in the ECASS I a positive effect of IV thrombolysis was revealed only after a careful retrospective review of the initial CT scans, with the application of strict, greater than one-third MCA territory hypodensity exclusion criteria, t°,3v This difference from the NINDS trial may be partly related to the longer, 6-hour treatment window applied to the ECASS patients. Furthermore, based on additional review of the ECASS data, it has been suggested that a subgroup of hyperacute stroke patients without demonstrable CT ischemia are also unlikely to benefit from IV thrombolysis. This is likely because, in patients with no or minimal ischemic change, the potential benefits of thrombolysis are outweighed by the risk of hemorrhage, to

Thus, the accurate, early identification of hypodense, ischemic parenchyma by CT has important clinical implications for acute stroke treatment. The advent of thrombolytic therapy makes the CT detection of hypodense brain tissue, as well as the better characterization of its predictive value for hemorrhage after treatment, exceedingly important. The ability of physicians to detect parenchy-

Fig 7. Noncontrast CT scans of 2 different patients showing hyperdense basilar artery signs indicating acute basilar thrombus. These were angiographically confirmed.


mal hypodensity, however, is limited by potentially clinically relevant inter- and intraobserver variability.l,9.42,44-46 In the first ECASS trial, nearly half of the protocol violations were because of failure to recognize early noncontrast CT of infarction. 1,4~ In discussing the ECASS, one editorialist pointed out that had the local, prospective readers been as able to detect subtle early signs of ischemia as the panel of 6 expert neuroradiologists who later retrospectively reviewed the ECASS scans, the study results would have been dramatically improved. 1

Some of the reported variability in CT scan interpretation is dependent on level of training. A study tested the ability of emergency physicians, neurologists, and general radiologists to determine eligibility for thrombolytic therapy from a group of acute head CYs. 46 Scan findings included intracerebral hemorrhage, acute infarction, intracerebral calcifications (a hemorrhage mimic), old cerebral infarction (an acute infarction mimic), and normal parenchyma. The average correct reads by all physicians for all CT scans was 77%. Of CT readings by emergency physicians, 67% were correct; of readings by neurologists, 83% were correct; and of readings by radiologists, 83% were correct. Overall sensitivity for detecting hemorrhage was 82%; 17% of emergency physicians, 40% of neurologists, and 52% of radiologists achieved 100% sensitivity for identification of hemorrhage. The investigators concluded that physicians in this study did not uniformly achieve a level of sensitivity for identification of intracerebral hemorrhage sufficient to permit safe selection of candidates for thrombolytic therapy.

Even brief periods of training, however, can improve interreader reliability in the early CT detection of stroke. At a 4-hour training course for participants in the second ECASS trial, readers were shown 2 different sets of 10 CT scans before and after the course. The scans shown included normal studies, studies with an acute less than one-third MCA territory stroke, and studies with an acute greater than one-third MCA territory stroke. After the completion of training, the average number of correct readings significantly increased? 8,49

Interestingly, among highly trained readers, an awareness of a patient's precise clinical history might not be an important factor in assessing subtle early CT signs of ischemia. In a 1996 study that included 12 normal and 33 abnormal CT scans

showing MCA infarction, unblinding the reviewers to the side of the infarction did not significantly change interobserver agreement or accuracy. It was concluded that, even without knowing the precise clinical history, neuroradiologists can assess subtle CT signs of hyperacute infarction with moderate to substantial interobserver agreement. 9 Of the CT findings of cortical swelling, HMCAS, and parenchymal hypodensity, the least interobserver variability was found for the detection of parenchymal hypodensity. However, the readers in this study were instructed to specifically look for acute stroke.

In a different study, when blinded readers were given no history or specific instructions to look for acute stroke, the sensitivity of acute stroke detection was 38%. The sensitivity of the same readers increased to 52% when the suspicion for acute stroke was noted in the available clinical history. 48

The Alberta Stroke Program Early CT Score has been proposed as a more reliable and reproducible method of evaluating acute stroke CT scans in comparison to the one-third rule. 5° The 10-point quantitative topographic CT scan score divides the MCA territory into 10 regions of interest. A point is subtracted for each region showing early ischemic changes. Scores of 7 or less are correlated with poor IV rt-PA outcomes. Interobserver and intraobserver reliability has been shown to be better with the Alberta Stroke Program Early CT Score than with the one-third rule. 49

There have been numerous reports comparing the sensitivity of CT with diffusion-weighted imaging in the diagnosis of acute stroke. These studies have universally shown the greater sensitivity of diffusion-weighted imaging. 5°44

SUMMARY

The advantages of CT scanning in the assess- ment of hyperacute stroke patients include conve- nience, accuracy, speed, and low cost. CT scanning is presently considered to be the standard of care for the detection of acute extra-axial and parenchy- real hemorrhage, although newer MR1 techniques are challenging this claim. 55-58 The accuracy of CT scanning for stroke detection can be optimized by the use of appropriate scanning technique, as well as interactive image review at a picture archiving and communication system workstation using narrow window and level settings. More importantly, CT scanning has prognostic value for patients

190 SYMONS ET AL

r ece iv ing t h r o m b o l y t i c t r e a tmen t by he lp ing to

p red ic t b o t h r e sponses to t r e a tmen t and h e m o r -

rhage risk. F ina l ly , as wil l be d i scussed in o ther

ar t ic les in this issue, C T a n g i o g r a p h y can be

p e r f o r m e d in the same i m ag i ng sess ion as noncon -

trast C T scann ing w i thou t subs tan t ia l ly inc reas ing

the t ime requ i red for pa t i en t eva lua t ion . This pro-

v ides data r ega rd ing vascu la r pa t ency and t i ssue

per fus ion , w h i c h are va luab le in the t r iage o f s t roke

pat ients .

REFERENCES

1. Russel E: Diagnosis of hyperacute ischemic infarct with CT: Key to improved clinical outcome after intravenous thrombolysis? Radiology 205:315-318, 1997

2. del Zoppo GJ: Acute stroke- on the threshold of a therapy. New Engl J Med 333:1632-1633, 1995

3. Powers WJ, Zivin J: Magnetic resonance imaging in acute stroke: Not ready for prime time. Neurology 50:842-843, 1998

4, Marks MP: CT in ischemic stroke. Neuroimaging Clin N Am 8:515-523, 1998

5. Brant-Zawadzki M: CT angiography in acute ischemic stroke: the right tool for the job? Am J Neuroradiol 18:1021- 1023, 1997

6. Dillon WP: CT techniques for detecting acute stroke and collateral circulation: in search of the Holy Grail. Am J Neuroradiol 19:191-192, 1998

7. Tomura N, Uemura K, Inugami A, et al: Early CT findings in cerebral infarction. Radiology 168:463-467, 1988

8. Truwit CL, Barkowich AJ, Gean-Marton A, et al: Loss of the insular ribbon: Another early CT sign of acute middle cerebral artery infarction. Radiology I76:801-806, 1990

9. von Kummer R, Holle R, Grzyska U, et al: Interobserver agreement in assessing early CT signs of middle cerebral artery infarction. Am J Neuroradiol 17:1743-1748, 1996

10. von Kummer R, Allen KL, Holle R, et al: Acute stroke: Usefulness of early CT findings before thrombolytic therapy. Radiology 205:327-333, 1997

11. Mohr J, Biller J, Hflal S, et al: Magnetic resonance versus computed tomographic imaging in acute stroke. Stroke 26:807-812, 1995

12. Gonzalez RG, Schaefer PW, Buonanno F, et al: Diffu- sion-weighted MR imaging: diagnostic accuracy in patients imaged within 6 hours of stroke symptom onset. Radiology 210:155-162, 1999

13. yon Kummer R, Meyding-Lamade U, Forsting M, et al: Sensitivity and prognostic value of early computed tomography in middle cerebral artery trunk occlusion. Am J Neuroradiol 15:9-15, 1994

14. Bell BA, Symon L, Branston NM: CBF and time thresholds for the formation of ischemic cerebral edema, and effect of reperfusion in baboons. J Neurosurg 62:31-41, 1985

15. Torack RM, Alcala H, Gado M, et al: Correlative assay of computerized cranial tomography (CTT) water content and specific gravity in normal and pathological postmortem brain. J Neuropathol Exp Neurol 35:385-392, 1976

16. Unger E, Littlefield J, Gado M: Water content and water structure in CT and MR signal changes: Possible influence in detection of early stroke. Am J Neuroradiol 17:1743-1748, 1988

17. Dix JA, Evans AJ, Kallmes DF, et ah Accuracy and precision of CT angiography in a model of the carotid artery bifurcation. Am J Neuroradiol 18:409-415, 1997

18. Lev MH, Ackerman RH, Chehade RJ, et al: Accurate spiral CT angiographic measurement of carotid artery luminal diameter: A phantom study. Radiology 201:117, 1996 (abstr)

19. Liu Y, Hopper KD, Mauger DT, et al: CT angiographic measurement of the carotid artery: optimizing visualization by manipulating window and level settings and contrast material attenuation. Radiology 217:494-500, 2000

20. Webb WR, Brant WE, Helms CA: Introduction to CT of the Abdomen and Pelvis. Fundamentals of Body CT. Philadel- phia, Saunders, 1991, p 137

21. Lev MH, Farkas J, Gemmete JJ, et al: Acute stroke: improved nonenhanced CT detection--Benefits of soft copy inteiqpretation by using variable window width and center level settings. Radiology 213:150-155, 1999

22. Woodward PK, Slone RM, Gierada DS, et al: Chest radiography: Depiction of normal anatomy and pathologic structures with selenium-based digital radiography versus conventional screen-film radiography. Radiology 203:196-201, 1997

23. Bryan S, Weatherburn G, Watkins J, et al: Radiology report times: Impact of picture archiving and communication systems. AIR 170:1153-1159, 1998

24. Pratt HM, Langlotz CP, Feingold ER, et al: Incremental cost of department-wide implementation of a picture archiving and communication system and computed radiography. Radi- ology 206:245-252, 1998

25. Bendszus M, Urbach H, Meyer B, et al: hnproved CT diagnosis of acute middle cerebral artery tenitory infarcts with density-difference analysis. Neuroradiol 39:127-131, 1997

26. Maldjian JA, Chalela J, Kasner SE, et aI: Automated CT segmentation and analysis for acute middle cerebral artery stroke. Am J Neuroradiol 22:1050-1055, 2001

27. von Kummer R, BouNuain H, Bastianello, et al: Early prediction of irreversible brain damage after ischemic stroke at CT. Radiology 219:95-100, 2001

28. Grond M, yon Kummer R, Sobesky J, et ah Early x-ray hypoattenuation of brain parenchyma indicates extended critical hypoperfusion in acute stroke. Stroke 31:133-139, 2000

29. Wall SD, Brant-Zawadski M, Jeffrey RB, et al: High frequency CT findings within 24 hours after cerebrai infarction. Am J Radiol 138:307-3110 1982

30. Bozzao L, Bastianello S, Fantozzi LM, et al. Correlation of angiographic and sequential CT findings in patients with evolving cerebral infarction. American Journal of Neuroradiol 1989; 10:1215-1222.

31. Pressman BD, Tourje EJ, Thompson JR. An early CT sign of ischemc infarction: increased density in a cerebral artery. Am J Neuroradiol 8:645-648, 1987

32. Koo CK, Teasdale E, Muir KW: What constitutes a true hyperdense middle cerebral artery sign? Cerebrovas Dis 10: 419-423, 2000


33. Barber PA, Demchuk AM, Hudon, et al: Hyperdense sylvian MCA "dot" sign: a CT marker of acute ischemia. Stroke 32:84-88, 2001

34. Leafy MG, Saver JL, Kidwell CS, et al: Validation of the CT MCA dot sign: an angiographic correlation study. In the Proceedings of the 27 th International Stroke Conference. Stroke 1:342, 2002 (abstr)

35. Tomsick TA, Brott T, Barsan W, et al: Thrombus localization with emergency cerebral computed tomography. Am J Neuroradiol 13:79-85, 1992

36. Adams HP, Brott TG, Ferlan AJ, et al: Guidelines for tbrombolytic therapy for acute stroke: A supplement to the guidelines for the management of patients with acute ischemic stroke. Circulation 94:1167-1174, 1996

37. National Institute of Neurological Disorders for Stroke rt-PA Study Group. Intracerebral hemorrhage after intravenous t-PA therapy for ischemic stroke. Stroke 28:2109-2118, 1997

38. Dubey N, Bakahi R, Wasay M, et al: Early computed tomography hypodensity predicts hemorrhage after intravenous tissue plasminogen activator in acute ischemic stroke. J Neuro- imaging 11:184-188, 2001

39. Larrne V, vorl Kummer RR, Muller A, et aI: Risk factors for severe hemorrhagic transformation in ischemic stroke patients treated with recombinant tissue plasminogen activator: A secondary analysis of the European-Australasian Acute Stroke Study (ECASS II). Stroke 32:438-441, 2001

40. Tomsick T, Brott T, Barsan W, Broderick J, et al. Prognostic value of the hyperdense middle cerebral artery sign and stroke scale score before ultraearly thrombolytic therapy. Am J Neuroradiol 17:79-85, 1996

41. Hacke W, Kaste M, Fieschi C, et al: Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke: The European Cooperative Acute Stroke Study (ECASS). JAMA 274:1017-1025, 1995

42. von Kummer R: Effect of training in reading CT scans on patient selection for ECASS II. Neurology 51:$50-52, 1998

43. Larrne V, von Kummer RR, Muller A, et aI: Risk factors for severe hemorrhagic transformation in ischemic stroke patients treated with recombinant tissue plasminogen activator: a secondary analysis of the European-Australasian Acute Stroke Study (ECASS II). Stroke 32:438-441, 2001

44. Tomsick T: Sensitivity and prognostic value of early CT in occlusion of the middle cerebral artery trunk: a commentary. Am J Neuroradiol 15:16-18, 1994

45. Shinar D, Gross CR, Hier DB, et al: Interobserver reliability in the interpretation of computed tomographic scans of stroke patients. Arch Neurol 44:149-155, 1987

46. Schriger DL, Kalafut M, Starkman S, et al: Cranial computed tomography interpretation in acute stroke: physician accuracy in determining eligibility for thrombolytic therapy [see comments]. JAMA 279:1293-1297, 1998

47. yon Kummer R, Holle R, Meier DH, et al: Can CT reading training enhance the safety of thrombolytic treatment in acute stroke? Proceedings of the 36th Annual Meeting of the American Society of Neuroradiol, Philadelphia, PA, 1998, pp 153-154

48. Mullins ME, Lev MH, Schellingerhout D, et al: Influ- ence of availability of clinical history on detection of early stroke when using unenhanced CT and diffusion-weighted MR imaging. Am J Radiol (in press)

49. Warwick Pexman JH, Barber PA, Hill MD et ai: Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. Am J Neuro- radiol 22:1534-1542, 2001

50. Urbach H, Flacke S, Keller E, et aI: Detectability and detection rate of acute cerebral hemisphere infarcts on CT and diffusion-weighted MRI. Neuroradiol 42:722-727, 2000

51. Lansberg MG, Albers GW, Beaulieu C, et al: Compari- son of diffusion-weighted MRI and CT in acute stroke. Neu- rology 54:1557-1561, 2000

52. Fiebach J, Jansen O, Schellinger P, et al: Comparison of CT with diffusion-weighted MRI in patients with hyperacute stroke. Neuroradiol 43:628-632, 2001AU

53. Schellinger PD, Fiebach JB, Jansen O, et al: Comparison of CT and diffusion-weighted imaging in acute stroke <6h: a randomized prospective trail. Proceedings of the 27 ~h Interna- tional Stroke Conference. Stroke 1:348, 2002 (abstr)

54. Barber PA, Darby DG, Desmond PM, et al: Identifica- tion of major ischemic change. Diffusion-weighted imaging versus computed tomography. Stroke 30:2059-2065, 1999

55. Singer MB, Atlas SW, Drayer BP: Subarachnoid space disease: diagnosis with fluid-attenuated inversion-recovery MR imaging and comparison with gadolinium-enhanced spin-echo MR imaging-blinded reader study. Radiology 208:417-422, 1998

56. Atlas SW, Thulborn KR: MR detection of hyperacute parerlchymal hemorrhage of the brain. Am J Neuroradiol 19:1471-1507, 1998

57. Gilleland GT, Gammal T, Fisher W: Hyperacute intra- ventricular hemorrhage revealed by gadolinium-enhanced MR imaging. Am J Radiol 170:787-789, 1998

58. Noguchi N, Ogawa T, Inugami A, et al: Acute subarachnoid hemorrhage: MR imaging with fluid attenuated inversion recovery MR imaging. Radiology 203:257-262, 1995

Documents

Noncontrast conventional computed tomography in the evaluation of acute stroke