Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2381041113
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, P. E.
Right arrow Articles by Frayne, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, P. E.
Right arrow Articles by Frayne, R.
(Radiology 2006;238:232-239.)
© RSNA, 2006

Neuroradiology

Acute Ischemic Stroke: Accuracy of Diffusion-weighted MR Imaging—Effects of b Value and Cerebrospinal Fluid Suppression1

Philip E. Chen, MD, Jessica E. Simon, MB ChB, Michael D. Hill, MD, Chul-Ho Sohn, MD2, Peter Dickhoff, MD, William F. Morrish, MD, Robert J. Sevick, MD and Richard Frayne, PhD

1 From the Seaman Family MR Research Centre (P.E.C., J.E.S., M.D.H., C.H.S., R.J.S., R.F.) and Calgary Stroke Program (J.E.S., M.D.H.), Foothills Medical Centre, Calgary Health Region, 1403 29th St NW, Calgary, AB, Canada T2N 2T9; and Departments of Clinical Neuroscience (J.E.S., M.D.H., R.J.S., R.F.), Radiology (C.H.S., P.D., W.F.M., R.J.S., R.F.), Medicine (M.D.H.), and Community Health Sciences (M.D.H.), University of Calgary, Calgary, Alberta, Canada. Received June 24, 2004; revision requested August 25; revision received February 1, 2005; accepted February 28. Supported by the Alberta Foundation for Health Research, Canadian Institutes for Health Research (CIHR), and the Heart and Stroke Foundation of Canada (HSFC). J.E.S. is an Alberta Heritage Foundation for Medical Research (AHFMR) Fellow; M.D.H. is an HSFC and CIHR Research Scholar; and R.F. is an AHFMR Medical Scholar, an HSFC Research Scholar, and a Canada Research Chair. Address correspondence to R.F. (e-mail: rfrayne{at}ucalgary.ca).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Purpose: To prospectively determine which diffusion-weighted magnetic resonance (MR) imaging technique (ie, conventional diffusion-weighted MR imaging [b = 1000 or 1500 sec/mm2] or fluid-inversion prepared diffusion [FLIPD] MR imaging [b = 1500 sec/mm2]) is most accurate in depicting acute ischemic stroke at 3 T.

Materials and Methods: The Health Research Ethics Board approved this study; written informed consent was provided by all participants or their surrogate. Diffusion-weighted MR imaging was performed in 75 consecutive patients (43 men, 32 women; mean age, 64.0 years) with acute ischemic stroke. Two experienced neuroradiologists determined the presence of hyperacute stroke lesions at diffusion-weighted MR imaging by locating areas of hyperintensity that corresponded to regions with a decreased diffusion coefficient. These findings were used as the reference standard. Four raters who were blinded to patient history assessed all images and apparent diffusion coefficient maps for the presence of changes that were consistent with acute ischemic stroke. Accuracy, sensitivity, specificity, negative predictive value, positive predictive value, and inter- and intrarater reliability scores were calculated for each technique.

Results: Specificity, positive predictive value, and accuracy were not significantly different among the techniques. FLIPD MR images obtained with a b value of 1500 sec/mm2 had decreased sensitivity for acute ischemic stroke (mean, 61.8%; 95% confidence interval [CI]: 55.4%, 67.9%) compared with conventional diffusion-weighted MR images obtained with a b value of either 1000 sec/mm2 (mean, 82.5%; 95% CI: 77.1%, 87.0%) or 1500 sec/mm2 (mean, 84.5%; 95% CI: 79.3%, 88.9%). FLIPD MR images also had decreased negative predictive value (mean, 96.5%; 95% CI: 95.7%, 97.2%) compared with conventional diffusion-weighted MR images obtained with a b value of either 1000 sec/mm2 (mean, 98.4%; 95% CI: 97.8%, 98.8%) or 1500 sec/mm2 (mean, 98.6%; 95% CI: 98.1%, 99.0%). Intra- and interrater reliability scores were generally excellent for all three techniques.

Conclusion: FLIPD MR images obtained with a b value of 1500 sec/mm2 are less suitable for the detection of acute ischemic stroke owing to a decreased sensitivity and negative predictive value. The performance of the two conventional diffusion-weighted MR imaging techniques (b = 1000 and 1500 sec/mm2) was equivalent.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Diffusion-weighted magnetic resonance (MR) imaging measures the random motion of water molecules in tissue. The increased need to assess acute ischemic stroke prior to thrombolysis has placed particular emphasis on the use of diffusion-weighted MR imaging (1). The observed signal intensity for diffusion-weighted MR imaging is calculated as e–D·b · e–TE/T2, where b is diffusion sensitivity, D is water diffusion, and TE is echo time, and T2 is T2 relaxation time. The observed signal intensity is influenced initially by the changes in water diffusion that are associated with cytotoxic edema secondary to ischemia and later by the T2 relaxation time that is associated principally with vasogenic edema (2). Because diffusion-weighted MR imaging is especially sensitive to shifts in water that occur between extracellular and intracellular compartments, this technique can demonstrate regions of brain that are undergoing ischemic injury within 1 hour or less of symptom onset. In patients with acute ischemic stroke, intracellular swelling, which occurs as a result of water shifting from the extracellular compartment, causes the initial increase in observed signal intensity. Other factors that contribute to the increase in signal intensity are the increased tortuosity of the extracellular and intracellular spaces and increased intracellular viscosity (3).

At diffusion-weighted MR imaging, image contrast can be controlled by varying diffusion sensitivity. This is accomplished by altering the strength of, duration of, and/or interval between diffusion gradients (4). Traditionally, a b value of 1000 sec/mm2 had been used to assess patients with acute ischemic stroke because the strength of the diffusion gradients was historically restricted by hardware performance limitations that made it difficult to achieve higher b values with acceptable echo times. Previous studies have sought to determine the optimal b value for detecting acute ischemic stroke. Kingsley and Monahan (5) stated that the contrast-to-noise ratio is optimal at lower b values, such as 1000 sec/mm2, although to our knowledge there are no studies that address whether a b value of 1000 sec/mm2 corresponds to improved detection of acute ischemic lesions. The use of higher b values may increase diffusion sensitivity by diminishing the hyperintensity of tissues with long T2 relaxation times (ie, T2 shine-through) and by improving the conspicuity of infarction in a hyperacute setting when only subtle diffusion abnormalities are present (6–8). High b values, however, may decrease absolute differences in signal intensity between ischemic and normal tissues, increase differentiation between the apparent diffusion coefficient (ADC) of gray matter and white matter, and produce hyperintense white matter tracts that may mask acute ischemic stroke lesions (9). This suggests that an intermediate b value may provide optimal visualization (6). Specifically, images obtained with a b value of 1500 sec/mm2 have been shown to have a higher contrast-to-noise ratio than those obtained with a standard b value of 1000 sec/mm2, though to the best of our knowledge this benefit has yet to be shown to equate with increased clinical detection of acute ischemic stroke (10).

Fluid-inversion prepared diffusion (FLIPD) MR imaging is a method that minimizes the areas of hyperintensity that result from fluids, such as cerebrospinal fluid (CSF), that have long T2 relaxation times. FLIPD MR imaging produces CSF-nulled diffusion-weighted images by coupling an inversion pulse to a conventional diffusion-weighted MR imaging sequence. FLIPD MR imaging has been shown to improve quantitative ADC measurements in regions of CSF (11) and has also been proposed as a potentially useful investigative tool in acute ischemic stroke (12). With CSF nulling, diffusion lesions that are located in regions with a long T2 relaxation time become more conspicuous (13). This is especially true for lesions located in the cortex, at the gray matter–white matter interface, and in periventricular regions, such as the insular cortex. Simon et al (14), however, found that acute ischemic stroke lesions were less conspicuous on FLIPD MR images and that these images demonstrated a decreased signal-to-noise ratio, thereby suggesting that this technique may not be suitable for the detection of acute ischemic stroke.

It is clinically important to assess which of these techniques is most effective in demonstrating acute ischemic stroke because the presence and size of ischemic lesions at diffusion-weighted MR imaging are important determinants of clinical treatment and prognosis (15). It is equally important to evaluate whether an increased contrast-to-noise ratio will translate into improved detection of acute stroke. Similar controversy exists with the use of FLIPD MR imaging in patients with acute ischemic stroke. On the basis of results from two previous studies that assessed contrast-to-noise ratio according to b value (10) and the use of FLIPD MR imaging (14), we hypothesized that a conventional diffusion-weighted MR imaging technique with b value of 1500 sec/mm2 would have the best performance in demonstrating acute ischemic stroke. Thus, the purpose of our study was to prospectively determine which diffusion-weighted MR imaging technique (ie, conventional diffusion-weighted MR imaging [b = 1000 or 1500 sec/mm2] or FLIPD MR imaging [b = 1500 sec/mm2]) is most accurate in depicting acute ischemic stroke at 3 T.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
A prospective cohort study was performed in 75 patients who had clinical stroke syndromes by using conventional diffusion-weighted MR imaging (b = 1000 sec/mm2 and 1500 sec/mm2) and FLIPD MR imaging (b = 1500 sec/mm2). Patients were enrolled from March 2001 to January 2003. All patients presented to the emergency department and were evaluated by the stroke neurology team. Included in this evaluation was a comprehensive neurologic examination. A qualified member of the stroke team determined the National Institutes of Health Stroke Scale (NIHSS) score at the time of MR imaging and 24 hours after symptom onset. Prior to MR imaging, all patients underwent diagnostic computed tomography (CT) and, if warranted, intravenous thrombolysis. Our Health Research Ethics Board approved this study, and written informed consent was provided by all participants or their surrogate.

A total of 75 patients (43 men, 32 women; average age, 64.0 years ± 14.8 [standard deviation]) were imaged within 6.4 hours ± 5.9 of symptom onset. At the time of MR imaging, the median NIHSS score was 4 (range, 0–29), and at 24 hours after symptom onset, the median NIHSS score was 2 (range, 0–21). Seventeen (23%) of 75 patients were given intravenous tissue plasminogen activator immediately prior to MR imaging. Study demographics are summarized in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1. Summary of Patient Demographics

 
MR Imaging
MR imaging was performed by using a 3-T imager (Signa; GE Healthcare, Waukesha, Wis) that was equipped with high-speed gradients (40 mT/m peak strength and 184-µsec rise time). All imaging was performed by using a standard quadrature head coil. Our acute stroke protocol included diffusion-weighted MR imaging, time-of-flight MR angiography, fluid-attenuated inversion-recovery MR imaging, and perfusion-weighted MR imaging. For this study, three diffusion-weighted MR imaging examinations were performed. Single-shot spin-echo diffusion-weighted MR imaging with 19 5-mm sections, 2-mm gaps, and a 32 x 19-cm field of view was performed by using a 192 x 192 acquisition matrix and 7000/96.5 (repetition time msec/echo time msec). This protocol allowed for high-resolution diffusion-weighted MR imaging of the entire head. Images were acquired with diffusion sensitivities of 1000 sec/mm2 and 1500 sec/mm2. FLIPD MR imaging (b = 1500 sec/mm2) was performed with an inversion pulse and an inversion time of 2250 msec and was followed by conventional diffusion-weighted MR imaging (b = 1500 sec/mm2). For FLIPD MR imaging (b = 1500 sec/mm2), repetition time was increased to 9000 msec, and thus 36 seconds was required to obtain each image. The conventional diffusion-weighted MR images required 28 seconds to obtain. About 15 seconds was required for completion of the preimaging before the start of each image acquisition. A b value of 1500 sec/mm2 was selected for FLIPD MR imaging because the results of previous studies have shown that this b value maximizes signal intensity differences in hyperacute stroke (10). Because it was the clinically required technique, diffusion-weighted MR imaging with a b value of 1000 sec/mm2 was performed first and was followed by diffusion-weighted MR imaging with a b value of 1500 sec/mm2 and FLIPD MR imaging (b = 1500 sec/mm2). All diffusion-weighted images were obtained during the same imaging session within 15 minutes of each other and were registered by using the same sagittal localizer image so that all images were implicitly coregistered. The acquired images were transferred to an offline workstation (Advantage Windows Workstation; GE Healthcare) for analysis by using manufacturer-provided software (Functool, version 2.5; GE Healthcare). ADC maps were constructed to help separate true acute infarction diffusion changes from T2 shine-through on the diffusion-weighted MR images.

Image Review
After data acquisition and processing, each of the 75 patients was associated with six sets of images obtained by using each of the three techniques. Each set of images included diffusion-weighted MR images and ADC maps and contained 19 sections for a total of 8550 images. A two-staged analysis was conducted on these data. First, two neuroradiologists with more than 6 years experience in acute ischemic stroke (C.H.S., R.J.S.) established by consensus the presence of lesions at conventional diffusion-weighted MR imaging in each hemisphere and at each section location for each patient. This initial assessment was followed by a review during which four new raters with more than 1 year of experience were blinded to patient history and were asked to identify the presence of lesions in each hemisphere for each section location, patient, and technique.

To define the presence of an acute ischemic stroke lesion, the two neuroradiologists were provided with the diffusion-weighted images for all three techniques, as well as with the corresponding ADC images. Custom-written display software was used to enable simultaneous location-by-location viewing of the three images and maps. The two neuroradiologists independently scored each hemisphere for each section location and patient (2850 observations). Each neuroradiologist looked for areas of hyperintensity on the diffusion-weighted MR images that corresponded to regions of decreased signal intensity on the ADC maps. Subsequently, discordant findings in 16 (1.1%) of 1425 total sections in 12 patients (four of whom had discordant findings in two adjacent sections) were resolved by consensus. These neuroradiologists also had access to the available follow-up CT (27 patients) and MR (52 patients) images. The findings of these neuroradiologists were assumed to be the reference standard for this study.

For the review by the four raters (J.E.S., M.H.D., P.D., W.F.M.), each set of matching diffusion-weighted MR images and ADC maps was assigned a random number. The randomly numbered image sets were then ordered and presented to the raters. Four experienced raters (one neuroradiologist with 5 years of experience, one neuroradiology fellow with 1 year of experience, one stroke neurologist with 5 years of experience, and one stroke neurology fellow with 2 years of experience) reviewed each image independently (ie, images acquired by using the three diffusion-weighted MR imaging techniques were reviewed separately and in a random order). A second custom-written program allowed the simultaneous review of the diffusion-weighted MR images and ADC maps. Each rater indicated the presence or absence of changes consistent with acute ischemic stroke at each section location and in each hemisphere for each technique and patient. To assess intrarater reliability, each of the four blinded raters was asked to reanalyze a set of 35 images at least 1 week after the initial evaluation. These 35 sets consisted of the first 12 sets of diffusion-weighted MR images obtained with a b value of 1000 sec/mm2, the first 12 sets of FLIPD MR images obtained with a b value of 1500 sec/mm2, and the first 11 sets of diffusion-weighted MR images obtained with a b value of 1500 sec/mm2.

Statistical Analysis
The results of the four raters were compared with the reference standard. True-positive values, true-negative values, and accuracy, as well as false-positive and false-negative values, were calculated for each rater and were pooled across all raters. Inter- and intrarater reliability scores were calculated by using unweighted {kappa} statistics. Subset analyses were performed by using logistic regression analysis to see if performance was affected by level of training (ie, staff vs fellow), medical specialty (ie, neuroradiologist vs stroke neurologist), NIHSS score, patient age, and time from symptom onset. All confidence intervals for proportions are exact binomial confidence intervals, and comparisons between proportions were made by using the Fisher exact test. Confidence intervals for {kappa} statistics were derived from the standard error as described by Fleiss et al (16). The {kappa} values were interpreted as suggested by Landis and Koch (17)—that is, 0.00–0.20, poor agreement; 0.20–0.40, some agreement; 0.40–0.60, moderate agreement; 0.60–0.80, good agreement; and 0.80–1.00, near perfect agreement. All tests were two-sided, and conventional levels of significance (P = .05) were used. A commercially available software program (Stata 8.0; Stata, College Station, Tex) was used for all analyses.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
By using the reference standard, it was determined that 50 (67%) of 75 patients had lesions that were consistent with acute ischemic stroke or transient ischemic attack. These findings corresponded to a prevalence of 8.5% (243 lesions in 2850 observations) for all hemispheres, section locations, and patients.

Figure 1 shows a representative set of images with corresponding diffusion-weighted MR images and ADC maps that was used to assess acute ischemic stroke. Additional images (ie, three-dimensional time-of-flight MR angiograms, fluid-attenuated inversion-recovery MR images, and perfusion-weighted images [eg, relative mean transit time maps]) are displayed for presentation completeness. Although this information was not available to the four raters, the information was available to the two neuroradiologists who established the reference standard. These diffusion-weighted MR images and ADC maps are consistent with the general findings of this study; the acute ischemic stroke lesion is visible on images acquired with all three diffusion-weighted MR imaging techniques and can also be seen on the corresponding ADC maps.



View larger version (128K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1e: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (171K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1f: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (160K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1g: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1h: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 


View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1i: Images of 67-year-old woman with untreated acute ischemic stroke 14 hours after symptom onset. (a) Three-dimensional time-of-flight MR angiogram (24/3.3; flip angle, 15°) demonstrates occluded left middle cerebral artery (arrow). (b) Fluid-attenuated inversion-recovery fast spin-echo MR image (9000/161/2250 [repetition time msec/echo time msec/inversion time msec]) demonstrates no changes. (c) Relative mean transit time map obtained with single-shot echo-planar gradient-echo MR imaging (2200/30; flip angle, 45°) during the infusion of 20 mL of MR contrast agent at 5 mL · sec–1 confirms occlusion in left middle cerebral artery (arrows). (d–f) Diffusion-weighted MR images and (g–i) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for d, e, g, and j and 9000/96.5/2250 for f and i) show changes consistent with acute ischemic stroke (curved arrow in d–i), including hyperintensity on diffusion-weighted images and corresponding hypointensity on ADC maps. All views are oblique transverse views that were obtained at the same section location except for a, which is a maximum intensity projection image through a slab centered on the circle of Willis.

 
For pooled results of the four raters (Table 2), the overall performance of each diffusion-weighted MR imaging technique was good. Only the sensitivity and negative predictive value of FLIPD MR imaging (b = 1500 sec/mm2) were significantly different; these values were found to be lower than those obtained with conventional diffusion-weighted MR imaging. Results were similar for each rater. Figure 2 demonstrates decreased sensitivity at diffusion-weighted MR imaging (b = 1500 sec/mm2) and FLIPD MR imaging (b = 1500 sec/mm2). In this patient, the decrease in signal-to-noise ratio led to a decrease in the conspicuity of the acute ischemic stroke lesion. In addition, the lesion conspicuity is decreased because of white matter anisotropy, which caused confounding hyperintensity on diffusion-weighted MR images and decreased ADC values at higher b values.


View this table:
[in this window]
[in a new window]

  		Table 2. Pooled Statistics for Sensitivity, Specificity, Positive Predicative Value, Negative Predicative Value, and Accuracy

 


View larger version (161K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Images of 73-year-old man with untreated minor stroke 17 hours after symptom onset. (a-c) Diffusion-weighted MR images and (d–f) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for a, b, d, and e and 9000/96.5 for c and f) demonstrate changes consistent with acute ischemia. Changes are clearly seen on images obtained with a b value of 1000 sec/mm2 (a, d) but are less well visualized on diffusion-weighted MR images obtained with a b value of 1500 sec/mm2 (b, e) or on FLIPD MR images obtained with a b value of 1500 sec/mm2 (c, f) owing to reduced signal-to-noise ratio. Increased diffusion-weighted MR imaging signal intensity and decreased ADC values on images obtained with a b value of 1500 sec/mm2 owing to white matter anisotropy also affected evaluation in this patient. Ischemic lesion can be seen on diffusion-weighted MR images and ADC maps (arrow in a–f). All views are oblique transverse views that were obtained at the same section location.

 


View larger version (175K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Images of 73-year-old man with untreated minor stroke 17 hours after symptom onset. (a-c) Diffusion-weighted MR images and (d–f) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for a, b, d, and e and 9000/96.5 for c and f) demonstrate changes consistent with acute ischemia. Changes are clearly seen on images obtained with a b value of 1000 sec/mm2 (a, d) but are less well visualized on diffusion-weighted MR images obtained with a b value of 1500 sec/mm2 (b, e) or on FLIPD MR images obtained with a b value of 1500 sec/mm2 (c, f) owing to reduced signal-to-noise ratio. Increased diffusion-weighted MR imaging signal intensity and decreased ADC values on images obtained with a b value of 1500 sec/mm2 owing to white matter anisotropy also affected evaluation in this patient. Ischemic lesion can be seen on diffusion-weighted MR images and ADC maps (arrow in a–f). All views are oblique transverse views that were obtained at the same section location.

 


View larger version (183K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: Images of 73-year-old man with untreated minor stroke 17 hours after symptom onset. (a-c) Diffusion-weighted MR images and (d–f) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for a, b, d, and e and 9000/96.5 for c and f) demonstrate changes consistent with acute ischemia. Changes are clearly seen on images obtained with a b value of 1000 sec/mm2 (a, d) but are less well visualized on diffusion-weighted MR images obtained with a b value of 1500 sec/mm2 (b, e) or on FLIPD MR images obtained with a b value of 1500 sec/mm2 (c, f) owing to reduced signal-to-noise ratio. Increased diffusion-weighted MR imaging signal intensity and decreased ADC values on images obtained with a b value of 1500 sec/mm2 owing to white matter anisotropy also affected evaluation in this patient. Ischemic lesion can be seen on diffusion-weighted MR images and ADC maps (arrow in a–f). All views are oblique transverse views that were obtained at the same section location.

 


View larger version (170K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d: Images of 73-year-old man with untreated minor stroke 17 hours after symptom onset. (a-c) Diffusion-weighted MR images and (d–f) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for a, b, d, and e and 9000/96.5 for c and f) demonstrate changes consistent with acute ischemia. Changes are clearly seen on images obtained with a b value of 1000 sec/mm2 (a, d) but are less well visualized on diffusion-weighted MR images obtained with a b value of 1500 sec/mm2 (b, e) or on FLIPD MR images obtained with a b value of 1500 sec/mm2 (c, f) owing to reduced signal-to-noise ratio. Increased diffusion-weighted MR imaging signal intensity and decreased ADC values on images obtained with a b value of 1500 sec/mm2 owing to white matter anisotropy also affected evaluation in this patient. Ischemic lesion can be seen on diffusion-weighted MR images and ADC maps (arrow in a–f). All views are oblique transverse views that were obtained at the same section location.

 


View larger version (163K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2e: Images of 73-year-old man with untreated minor stroke 17 hours after symptom onset. (a-c) Diffusion-weighted MR images and (d–f) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for a, b, d, and e and 9000/96.5 for c and f) demonstrate changes consistent with acute ischemia. Changes are clearly seen on images obtained with a b value of 1000 sec/mm2 (a, d) but are less well visualized on diffusion-weighted MR images obtained with a b value of 1500 sec/mm2 (b, e) or on FLIPD MR images obtained with a b value of 1500 sec/mm2 (c, f) owing to reduced signal-to-noise ratio. Increased diffusion-weighted MR imaging signal intensity and decreased ADC values on images obtained with a b value of 1500 sec/mm2 owing to white matter anisotropy also affected evaluation in this patient. Ischemic lesion can be seen on diffusion-weighted MR images and ADC maps (arrow in a–f). All views are oblique transverse views that were obtained at the same section location.

 


View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2f: Images of 73-year-old man with untreated minor stroke 17 hours after symptom onset. (a-c) Diffusion-weighted MR images and (d–f) ADC maps obtained with single-shot echo-planar spin-echo MR imaging (7000/96.5 for a, b, d, and e and 9000/96.5 for c and f) demonstrate changes consistent with acute ischemia. Changes are clearly seen on images obtained with a b value of 1000 sec/mm2 (a, d) but are less well visualized on diffusion-weighted MR images obtained with a b value of 1500 sec/mm2 (b, e) or on FLIPD MR images obtained with a b value of 1500 sec/mm2 (c, f) owing to reduced signal-to-noise ratio. Increased diffusion-weighted MR imaging signal intensity and decreased ADC values on images obtained with a b value of 1500 sec/mm2 owing to white matter anisotropy also affected evaluation in this patient. Ischemic lesion can be seen on diffusion-weighted MR images and ADC maps (arrow in a–f). All views are oblique transverse views that were obtained at the same section location.

 
Specificity, sensitivity, negative predictive value, positive predictive value, and accuracy were calculated according to level of training, medical specialty, NIHSS score, patient age, and time from symptom onset. No significant differences were observed for the level of training (fellow vs staff, P > .05) or medical specialty (neuroradiology vs stroke neurology, P > .05). A slight effect was observed for NIHSS score such that, at higher NIHSS scores, lesions were more likely to be seen on conventional diffusion-weighted MR images obtained with a b value of 1000 sec/mm2. The percentage of image sections that demonstrated a lesion ranged monotonically from 7% with NIHSS scores between 0 and 4 to 15% with NIHSS scores of 24 or higher (P = .03). This difference was not related to the various diffusion-weighted MR imaging techniques (P = .790). Similarly, a nonsignificant age effect was seen that was independent of technique—that is, a lesion was more likely to be present in older patients (eg, at diffusion-weighted MR imaging [b = 1000 sec/mm2], 8% of patients younger than 50 years and 11% of patients older than 80 years demonstrated a lesion [P = .136]). No statistically significant differences were found between the time from symptom onset and technique or lesion presence at diffusion-weighted MR imaging (P = .475).

For the pooled intrarater and interrater reliability scores (Table 3), the reliability was generally excellent for all three techniques; however, the intrarater reliability for FLIPD MR imaging was significantly lower (P > .05) than that of conventional diffusion-weighted MR imaging. Also, compared with the other two techniques, conventional diffusion-weighted MR imaging (b = 1500 sec/mm2) had a lower interrater reliability.


View this table:
[in this window]
[in a new window]

  		Table 3. Pooled {kappa} Statistics for Inter- and Intrarater Reliability

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Our results demonstrate that while conventional diffusion-weighted MR images obtained with a b value of either 1000 or 1500 sec/mm2 provide essentially similar diagnostic interpretations, FLIPD MR images obtained with a b value of 1500 sec/mm2 are less sensitive and have a lower negative predictive value. This result is consistent with findings from prior studies that suggest that the decreased signal-to-noise ratio on FLIPD MR images can translate to decreased lesion conspicuity (14). Thus, a subsequent increase in false-negative assessments can occur, as well as a possible underestimation of lesion size (14). Additionally, the decreased level of hyperintensity in structures that are adjacent to CSF on FLIPD MR images did not improve rater sensitivity; in fact, the lower sensitivity exhibited at FLIPD MR imaging may be attributed to raters discounting true acute ischemic stroke changes as CSF-nulling artifacts. It is possible that the sensitivity and negative predictive value of FLIPD MR imaging were predominantly lowered in specific regions of the brain (ie, regions bordering CSF). Results of prior research suggest that FLIPD MR imaging may still be useful as an accurate technique in estimating the extent of reversible ischemic injury (12); however, the results of prior studies have also shown diffusion-weighted MR imaging to be a reliable marker of irreversible infarction (8,18–20). Although FLIPD MR imaging may still be useful as a prognostic indicator, particularly when quantifying ADC values (11), our findings suggest that its utility in acute ischemic stroke detection is limited.

Surprisingly, the previously documented increase in signal intensity for acute ischemic stroke changes that were associated with a b value of 1500 sec/mm2 (10) did not lead to a significant improvement in acute ischemic stroke detection. Yoshiura et al (21) noted that, at diffusion-weighted MR imaging, the contrast between normal gray and white cerebral tissues increased at higher b values. Accordingly, a lower b value (eg, 1000 sec/mm2) would provide a more uniform background, which might make more subtle lesions easier to detect. With increasing b value, white matter decreases in signal intensity less than does gray matter, which causes relative hypointensity in cortical gray matter and hyperintensity in white matter. This might lead to a loss of normal anatomic surface landmarks in the brain, difficulty in distinguishing acute ischemic stroke lesions from hyperintense white matter, or difficulty in detecting acute ischemic stroke lesions near the gray matter–white matter border (8).

The inter- and intrarater reliability scores were all good to excellent (22). The slightly lower interrater reliability for diffusion-weighted MR imaging (b = 1500 sec/mm2) and FLIPD MR imaging (b = 1500 sec/mm2) are possibly the reflection of an experience bias against these modalities, which are usually not included in the standard clinical protocol for acute ischemic stroke. Lack of exposure to these modalities could limit our ability to benefit from the increased diffusion sensitivity available with a b value of 1500 sec/mm2 or to become familiar with the expected artifacts of FLIPD MR imaging. Experience bias may also explain the slightly higher intrarater reliability score for diffusion-weighted MR images obtained with a b value of 1000 sec/mm2 ({kappa} = 0.85). Intrarater reliability scores for diffusion-weighted MR imaging (b = 1500 sec/mm2) and FLIPD MR imaging (b = 1500 sec/mm2) were 0.81 and 0.83, respectively. Intrarater variability was slightly worse for FLIPD MR imaging, perhaps owing to the experience bias.

For all three techniques, no significant differences in acute ischemic stroke detection were encountered with respect to patient age, NIHSS score at the time of imaging, time from symptom onset to MR imaging, medical specialty of the observer (neuroradiologist or stroke neurologist), or level of training (staff or fellow).

Limitations of this study included the use of a 3-T imager, the heterogeneity of ischemic strokes studied, the lack of randomization in the diffusion-weighted MR imaging acquisitions, and an inability to generate confidence intervals for intraobserver variability. For most acute ischemic stroke studies, researchers have used a 1.5-T MR imager; in our study, we used a 3-T MR imager. Theoretically, diffusion-weighted MR imaging properties (eg, b-value diffusion sensitivity) are not affected by field strength (14). Likewise, fluid-attenuated inversion-recovery MR imaging with an inversion time of 2250 msec produces fluid-nulled images at both 1.5 T and 3 T (22). These findings suggest that our results at 3 T would be similar to those at 1.5 T. The median NIHSS score was 4, which suggests that our study group experienced relatively minor strokes with perhaps less evolved ischemic changes. In our study group, 23% of the patients also underwent thrombolysis, most immediately prior to MR imaging, and some of these patients may have experienced recanalization before or during MR imaging. The inability to randomize the acquisition order of the three techniques (owing to ethical constraints) is potentially problematic; however, the short interval during which all three techniques were performed (<15 minutes) greatly minimizes any bias. The study design (ie, evaluating the presence of a lesion on the basis of hemisphere and section location) was a practical decision; however, such a design prevented us from compensating for correlations owing to lesions that are located in more than one section. As a result, it was not possible to determine confidence intervals for the intraobserver variability, though a clinically insignificant range of {kappa} values was observed ({kappa} = 0.78–0.86).

In conclusion, when compared with the FLIPD MR images (b = 1500 sec/mm2), conventional diffusion-weighted MR images obtained with a b value of 1000 or 1500 sec/mm2 were found to be the most accurate for investigating and detecting acute ischemic stroke. The FLIPD MR imaging technique was found to have a decreased sensitivity and decreased negative predictive value—both of which are critical factors in determining treatment options and patient prognosis. Diffusion-weighted MR images obtained with a b value of 1000 or 1500 sec/mm2 were equivalent in accuracy relative to the reference standard. The interrater reliability of conventional images obtained with a b value of 1500 sec/mm2 was less than that of images obtained with a b value of 1000 sec/mm2.


   ACKNOWLEDGMENTS
 
The support of the Calgary Stroke Program in providing access to patients who were imaged in this study and the technical assistance of Brian O'Brien in writing the anonymous display software used in this study are acknowledged.


   FOOTNOTES
 

Abbreviations: ADC = apparent diffusion coefficient • CSF = cerebrospinal fluid • FLIPD = fluid-inversion prepared diffusion • NIHSS = National Institutes of Health Stroke Scale

2 Current address: Department of Radiology, Keimyung University School of Medicine, Daegu, Korea. Back

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, all authors; study concepts, P.E.C., J.E.S., M.D.H., R.J.S., R.F.; study design, P.E.C., J.E.S., M.D.H., R.J.S., R.F., C.H.S.; literature research, P.E.C., R.F.; clinical studies, J.E.S., M.D.H., W.F.M., P.D., C.H.S., R.J.S.; data acquisition, all authors; data analysis/interpretation, P.E.C., J.E.S., M.D.H., R.F.; statistical analysis, P.E.C., M.D.H., R.F.; manuscript preparation, P.E.C., J.E.S., M.D.H., R.F.; manuscript definition of intellectual content, all authors; manuscript editing, P.E.C.; manuscript revision/review, P.E.C., M.D.H., R.F.; manuscript final version approval, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Beauchamp NJ Jr, Barker PB, Wang PY, vanZijl PC. Imaging of acute cerebral ischemia. Radiology 1999;212:307–324.[Abstract/Free Full Text]
  2. van Everdingen KJ, van der Grond J, Kappelle LJ, Ramos LM, Mali WP. Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke 1998;29:1783–1790.[Abstract/Free Full Text]
  3. Schaefer PW, Grant PE, Gonzalez RG. Diffusion-weighted MR imaging of the brain. Radiology 2000;217:331–345.[Abstract/Free Full Text]
  4. Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986;161:401–407.[Abstract/Free Full Text]
  5. Kingsley PB, Monahan WG. Selection of the optimum b factor for diffusion-weighted magnetic resonance imaging assessment of ischemic stroke. Magn Reson Med 2004;51:996–1001.[CrossRef][Medline]
  6. DeLano MC, Cao Y. High b-value diffusion imaging. Neuroimaging Clin N Am 2002;12:21–34.[CrossRef][Medline]
  7. Meyer JR, Gutierrez A, Mock B, Hebron D, Prager JM, Gorey MT, Homer D. High-b-value diffusion-weighted MR imaging of suspected brain infarction. AJNR Am J Neuroradiol 2000;21:1821–1829.[Abstract/Free Full Text]
  8. Burdette JH, Elster AD. Diffusion-weighted imaging of cerebral infarctions: are higher b values better? J Comput Assist Tomogr 2002;26:622–627.[CrossRef][Medline]
  9. Thierry A, Huisman GM. Diffusion-weighted imaging: basic concepts and application in cerebral stroke and head trauma. Eur Radiol 2003;13:2283–2297.[CrossRef][Medline]
  10. Pereira RS, Harris AD, Sevick RJ, Frayne R. Effect of b value on contrast during diffusion-weighted magnetic resonance imaging assessment of acute ischemic stroke. J Magn Reson Imaging 2002;15:591–596.[CrossRef][Medline]
  11. Latour LL, Warach S. Cerebral spinal fluid contamination of the measurement of the apparent diffusion coefficient of water in acute stroke. Magn Reson Med 2002;48:478–486.[CrossRef][Medline]
  12. Bykowski JL, Latour LL, Warach S. More accurate identification of reversible ischemic injury in human stroke by cerebrospinal fluid suppressed diffusion-weighted imaging. Stroke 2004;35:1100–1106.[Abstract/Free Full Text]
  13. Kamran S. Significance of hyperintense vessels on FLIPD MRI in acute stroke. Neurology 2000;55:265–269.[Abstract/Free Full Text]
  14. Simon J, Czechowsky DK, Hill MD, Harris AD, Buchan AM, Frayne R. FLAIR prepared diffusion-weighted imaging is not an improvement over conventional diffusion-weighted imaging in hyperacute stroke. AJNR Am J Neuroradiol 2004;25:1653–1658.[Medline]
  15. Lovblad KO, Baird AE, Schlaug G, et al. Ischemic lesion volumes in acute stroke by diffusion-weighted magnetic resonance imaging correlate with clinical outcome. Ann Neurol 1997;42:164–170.[CrossRef][Medline]
  16. Fleiss JL, Cohen J, Everitt BS. Large sample standard errors of kappa and weighted kappa. Psychol Bull 1969;72:323–327.[CrossRef]
  17. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1974;33:159–174.
  18. Moseley ME, Kucharczyk J, Mintorovitch J, et al. Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility-enhanced MR imaging in cats. AJNR Am J Neuroradiol 1990;11:423–429.[Abstract]
  19. Warach S, Dashe JF, Edelman R. Clinical outcome in ischemic stroke predicted by early diffusion-weighted and perfusion magnetic resonance imaging: a preliminary analysis. J Cereb Blood Flow Metab 1996;16:53–59.[CrossRef][Medline]
  20. Sorensen AG, Buonanno FS, Gonzalez RG, et al. Hyperacute stroke: evaluation with combined multisection diffusion-weighted and hemodynamically weighted echo-planar MR imaging. Radiology 1996;199:391–401.[Abstract/Free Full Text]
  21. Yoshiura T, Wu O, Zaheer A, Reese TG, Sorenson AG. Highly diffusion-sensitized MRI of brain: dissociation of gray and white matter. Magn Reson Med 2001;45:734–740.[CrossRef][Medline]
  22. Frayne R, Goodyear BG, Dickoff P, Lauzon ML, Sevick RJ. Magnetic resonance imaging at 3.0 T: technical challenges and benefits in neurological imaging. Invest Radiol 2003;38:385–402.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Am. J. Neuroradiol.Home page
P. Mukherjee, J.I. Berman, S.W. Chung, C.P. Hess, and R.G. Henry
Diffusion Tensor MR Imaging and Fiber Tractography: Theoretic Underpinnings
AJNR Am. J. Neuroradiol., April 1, 2008; 29(4): 632 - 641.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, P. E.
Right arrow Articles by Frayne, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chen, P. E.
Right arrow Articles by Frayne, R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE