FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2100064617> ?p ?o ?g. }

• W2100064617 endingPage "200" @default.
• W2100064617 startingPage "193" @default.
• W2100064617 abstract "AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract Purpose To compare isotropic (combined diffusion-weighted image [CMB], apparent diffusion coefficient [ADC], TRACE, exponential ADC [eADC], and isotropically-weighted diffusion image [isoDWI]) and anisotropic (relative anisotropy [RA], fractional anisotropy [FA], and volume ratio [VR]) diffusion images collected with fast magnetic resonance (MR) diffusion-weighted (DWI) and diffusion-tensor (DTI) acquisition strategies (each less than one minute) in hyper-acute stroke. Materials and Methods Twenty-one patients suffering from ischemic stroke—imaged within six hours of symptom onset using both DWI and DTI—were analyzed. Regions of interest were placed in the ischemic lesion and in normal contralateral tissue and the percent difference in image intensity was calculated for all nine generated images. Results The average absolute percent changes for the isotropic strategies were all > 38%, with isoDWI found to have a difference of 50.7% ± 7.9% (mean ± standard error, P < 0.001). The ADC maps had the most significant difference (–42.4% ± 2.0%, P < 0.001, coefficient of variation = 0.22). No anisotropic images had significant differences. Conclusion Anisotropic maps do not consistently show changes in the first six hours of ischemic stroke; therefore, isotropic maps, such as those obtained using DWI, are more appropriate for detecting hyper-acute stroke. Anisotropic images, however, may be useful to differentiate hyper-acute stroke from acute and sub-acute stroke. J. Magn. Reson. Imaging 2004;20:193–200. © 2004 Wiley-Liss, Inc. DIFFUSION-SENSITIVE IMAGING is an indicator of acute cerebral infarction, with animal studies showing image changes within minutes of vessel occlusion (1). Diffusion-weighted imaging (DWI) and diffusion-tensor imaging (DTI) are two magnetic resonance imaging (MRI) techniques that characterize water diffusion and are useful for detecting and assessing ischemic stroke (2–14). In this study, we report on which diffusion post-processed image produces the best diffusion images (or maps) of hyper-acute cerebral infarct in terms of image contrast between the ischemic and contralateral normal regions. Generally, there are two classes of images that can be generated from diffusion acquisitions: 1) those that produce isotropic (or directionally independent) maps, and 2) those that produce anisotropic (or directionally dependent) maps. Typically, in DWI, diffusion gradients are sensitized to diffusion along three orthogonal axes. The generation of the combined diffusion-weighted (CMB) images and the apparent diffusion coefficient (ADC) maps removes any directional effects, thus resulting in isotropic images. A number of DWI studies have used CMB images and ADC maps during the hyper-acute phase of human stroke (i.e., within six hours of symptom onset) and confirmed that these maps are sensitive to identifying infarct (3, 5, 8, 14). More recently, DTI acquisitions that characterize diffusion as a 3 × 3 tensor have been developed and used to study stroke. Because the diffusion tensor enables eigenvalue-based characterization for directionally dependent, rotationally-invariant diffusion, anisotropic images can be constructed in addition to isotropic images. Anisotropic diffusion imaging appears also to consistently show increased changes between the infarct and the contralateral normal tissue in acute (<24 hours after onset) and subacute (three to five days) populations (13). Other studies (4, 6, 7, 9-11, 15) have suggested that anisotropic changes might also show early diffusion effects due to ischemia, although none of these anisotropic studies solely studied hyper-acute patients (i.e., within six hours of stroke onset). Furthermore, some investigators used a DTI technique with a long scan time (10, 11) and/or employing multiple signal averages or sub-sampling (collecting more than six directionally different samples) (16) of the diffusion tensor (7). Both are unlikely to be acceptable in a rapid acute stroke protocol (14). No study to date has measured the clinical utility of anisotropic images in hyper-acute stroke. The purpose of this study was to compare the image contrast between infarcted and normal tissues when using various isotropic and anisotropic post-processing strategies in hyper-acute human ischemic stroke. As such, this study is unique and within the time frame relevant to treatment of human stroke. Short DWI and DTI acquisitions sequences (less than one minute for each) were used to reflect constraints on imaging in the hyper-acute setting (14). Specifically, in this study we were interested to know whether anisotropic maps acquired from the DTI data have any advantage in detecting hyper-acute ischemic stroke over isotropic maps. Our specific hypothesis was that both the isotropic and anisotropic images will show changes within six hours of stroke. MATERIALS AND METHODS This study examined 31 consecutive patients who presented to our institution with hyper-acute ischemic stroke (less than six hours from onset of symptoms) and who were subsequently entered into ongoing serial stroke imaging research studies. All protocols were approved by our Institutional Review Board and written informed consent was obtained before patient participation. Each patient was assessed by a stroke neurologist and received an emergent computed tomography (CT) study to rule out hemorrhage and assess early signs of ischemia (17). If appropriate, patients then received intravenous thrombolytic therapy (18). At the time of MRI, all patients were in the hyper-acute phase of infarction and the time from last known to be normal was 5.5 hours or less. National Institutes of Health Stroke Scale (NIHSS) scores (19) were obtained on all patients at time of imaging and, in all but one patient, at 24 hours after onset. Image Acquisition DWI acquires four images per slice location and, depending on the reconstruction strategy selected, may produce up to five images per location. Data is acquired with no diffusion sensitivity and with sensitivity to diffusion along each orthogonal axis, i.e., Dxx, Dyy and Dzz (2, 3, 5, 8, 10). These data are used to form a CMB and a T2-weighted (T2-w) image (see Fig. 1) and, optionally, diffusion images for each direction. DTI acquires more data and this allows the complete characterization of the second-order diffusion tensor (20). The simplest DTI acquisition acquires seven images (21, 22); one image without diffusion weighting, and six images sensitive to different components of the diffusion tensor (Dxx, Dyy, Dzz, Dxy, Dxz, Dyz). From this data, the nine elements of the 3 × 3 symmetric diffusion tensor can be obtained (i.e., Dyx = Dxy). By determining the complete diffusion tensor, the eigen values and eigen vectors can be calculated to generate rotationally-invariant representations of anisotropic diffusion (in the form of a diffusion ellipsoid). This processing may provide additional information about the micro-structural properties of tissue (23). Figure 1Open in figure viewerPowerPoint Illustration of DWI and DTI post-processing strategies examined in this study. T2-w image (a), CMB diffusion image (b), ADC map (c), TRACE of diffusion tensor (d), eADC map (e), isoDWI map (f), RA map (g), FA map (h), and VR map (i). Shown are the principle steps in the data flow starting with 1) the acquired diffusion data (left column); 2) calculation of the eigen values (λ1, λ2, and λ3) of the diffusion tensor (2nd column), if required; 3) generation of a diffusion index from the tensor or eigen values (3rd column); and 4) creation of a parametric map (right column). The T2-w and CMB images are formed from DWI data (only Dx, Dy, and Dz values are acquired). The ADC map is derived from the CMB image. In our study, the TRACE, eADC, and isoDWI maps were formed from the DTI data, although only the elements on the main diagonal are used. The RA, FA, and VR maps are formed from the eigen values. D is the mean of the diagonal values, or trace/3. Images shown (from patient 8) in this example were acquired 2.0 hours after symptom onset. Baseline NIHSS was 4 and 24 hours; NIHSS was 2. Infarct region (white arrow in b) was located in GM. All subjects were scanned on a 3.0-T MR scanner (Signa; General Electric, Waukesha, WI) equipped with high-performance gradients (40 mT m–1 peak strength, 150 mT m–1 msec–1 slew rate) and a transmit-receive quadrature head coil. Essentially the same single-shot echo-planar imaging technique was used for both DWI and DTI image acquisition, with a diffusion sensitivity of b = 1000 seconds mm–2. Nineteen 5-mm oblique-axial slices were acquired with 2-mm inter-slice gap using a field of view (FOV) of 32 cm × 19.2 cm. A 192 × 115 acquisition matrix was acquired and reconstructed to form a rectangular 256 × 153 image. The reconstructed voxel sizes for the DWI and DTI images were 1.25 × 1.25 × 5 mm. For DWI, the TR/TE (repetition time/echo time) was 7000 msec/99 msec, and for DTI, 7000 msec/105 msec. The small differences in TE are due to the subtle implementation differences between the optimized clinical DWI and DTI research sequences available at the time of the study. Total acquisition time for DWI was 28 seconds and for DTI was 49 seconds. For DTI, the gradient encoding scheme described by Basser and Pierpaoli (22) was used. No signal averaging was performed because it was our purpose to evaluate the post-processing of data collected using fast sequences appropriate for hyper-acute stroke imaging (14). DWI is a part of our MR acute stroke protocol (12), and DTI was preformed on these patients in addition to this routine protocol. Post-Processing Strategies Figure 1 illustrates how the acquired DWI and DTI data can be used to calculate parametric maps. In this study, isotropic ADC maps were calculated from the DWI image data. The DTI data was used to calculate both isotropic and anisotropic maps, including the trace of the tensor (sum of the main diagonal elements, TRACE). In principle, TRACE values are directly related to the ADC values (i.e., the ADC is equivalent to one third of the TRACE) (13, 22). TRACE was calculated after decoding the acquired linear combinations of the diffusion tensor encodings (22), but before eigen decompositions. This process would alter noise characteristics, although it is noted that in an ideal, noise-free environment, the ADC and TRACE images should be scaled versions of each other. Isotropically diffusion-weighted image (isoDWI) (9, 10) is similar to the CMB map but is TRACE-weighted rather than weighted by the ADC. While similar to isoDWI, exponential ADC (eADC) (9) maps use the TRACE-weighted diffusion exponent, but unlike isoDWI maps, they do not have any T2-weighting. The relative anisotropy (RA) map is the ratio of the magnitudes of the anisotropic and isotropic parts of the diffusion tensor (4, 22, 23). The fractional anisotropy (FA) map is the ratio of the anisotropic component of the diffusion tensor to the whole tensor, and the volume ratio (VR) map is the ratio of the volume of the diffusion ellipsoid with the volume of an equivalent isotropic diffusion sphere (4, 13). The images and maps assessed in this study were 1) automatically generated by the scanner reconstruction software (CMB, T2-w), 2) produced using commercially available software (Functool; General Electric Medical Systems) (ADC, TRACE, eADC, isoDWI), or 3) calculated using custom software, written in a high-level image processing language (IDL; Research Systems Inc., Boulder, CO), to first obtain the diffusion tensor and the eigen values of the diffusion tensor (λ1, λ2, and λ3) before calculating the anisotropic indices (RA, FA, VR). Data and Statistical Analysis Because reduced diffusion (as seen as a hyperintensity on CMB images and a hypointensity on an ADC map) is widely recognized as acute and hyper-acute stroke regions, these images and maps were used for the identification of the stroke region. A region of interest (ROI) was placed in the central 25% of the infarcted region (by area) so as to exclude any non-infarcted tissue and only include the most severely ischemic tissue. A second ROI, having the same area, was placed in a contralateral area (normal tissue), which in this study served as the control region for determining image intensity changes. These regions were propagated to the other maps and the normalized percent difference between infarcted and normal tissues was calculated on each image or map. In order to assess intra-operator variability, the placement of the ROIs was repeated four times in each patient. The location of the infarct ROI was determined by a stroke neurologist according to anatomic location (predominately white matter [WM], predominately grey matter [GM], or in a region of approximately equal white matter and grey matter [BOTH]), and the location identification was verified and confirmed by a second stroke neurologist and a neuroradiologist, both of whom were blinded to the initial assessment. Analysis of variance (ANOVA) was used to assess the percent difference between infarcted and normal regions among the different post-processing techniques. First, intra-observer variability was assessed as a repeated variable within a repeated measures ANOVA. Upon confirmation of no intra-observer variability, the average of the four measures was used for subsequent analyses. A univariate ANOVA was used to test for an interaction between the infarct location and image technique. Finally, using an ANOVA, we compared the mean percent difference for each post-processing technique to the mean percent difference for T2-w imaging. Follow-up planned comparisons were used to determine the source of the interaction. As an exploratory analysis, we used the same ANOVA model, dividing infarcts according to tissue-type, and compared the signal intensity of the mean percent difference at each location to the T2-w technique. Statistical significance was defined at a level of α = 0.05 for the omnibus analyses and the adjusted level of α = 0.006 based upon the Bonferroni error correction for the planned comparisons. RESULTS Thirty-one patients met the inclusion criteria and were enrolled in this study. Ten patients were subsequently excluded due to either 1) the absence of a focal diffusion-weighted lesion (seven patients), or 2) significant image artifact arising from either technical difficulties and/or patient movement (three patients). Demographic information for the remaining 21 patients is summarized in Table 1. This group had 16 men and five women with a mean age of 70.1 years ± 2.4 years (mean ± standard error). The mean stroke onset-to-MR scan time was 3.7 hours ± 0.2 hours. At time of imaging, the median NIHSS score was 4 with a inter-quartile range of 3–10. At 24 hours, median NIHSS was 3.5 with an inter-quartile range of 2–7.25. Of these 21 patients, five (24%) were treated with intravenous tissue plasminogen activator (tPA; Activase; Genetech, San Francisco, CA) within three hours of onset, immediately before or during MRI. Table 1. Patient Demographics, Treatment, and Infarct Location of the Study Group Patient number Age M/F Delay to scan (h) Baseline NIHSS score 24 h NIHSS score Treated with tPA Infarct location 1 66 M 3.0 1 1 No WM 2 83 M 3.7 10 13 Yes GM 3 73 F 2.1 1 3 No WM 4 65 M 2.9 3 3 No Both 5 62 M 5.5 15 12 No WM 6 71 F 4.2 20 5 Yes WM 7 70 M 5.0 3 2 No GM 8 60 M 2.0 4 2 No GM 9 55 F 3.4 9 0 No WM 10 79 M 4.2 6 7 No GM 11 62 M 5.0 16 25 No Both 12 81 F 2.7 15 N/A Yes WM 13 83 M 3.6 6 6 No WM 14 48 M 3.6 2 8 No Both 15 75 M 4.8 3 5 No Both 16 91 M 5.5 26 11 Yes WM 17 82 F 2.9 2 0 No WM 18 73 M 4.3 3 4 No WM 19 69 M 3.7 6 2 Yes Both 20 60 M 2.2 3 3 No Both 21 64 M 2.6 3 2 No Both Mean 70.1 3.7 SD 10.7 1.1 Median 70 3.6 4 3.5 Inter-quartile range 3–10 2–7.25 M = male, F = female, NIHSS = National Institutes of Health Stroke Scale, tPA = tissue plasminogen activator, SD = standard deviation. Figure 1 presents a GM stroke imaged 2.0 hours after stroke onset (patient 8 in Table 1). A region of infarct is clearly seen on the CMB, ADC, TRACE, eADC, and isoDWI images and maps. No consistent changes were observed on the T2-w images or the RA, FA, and VR maps. Figure 2 shows the nine images/parametric maps acquired or generated in a second patient 5.5 hours after onset (patient 5 in Table 1). A region of infarct in WM is clearly seen on the CMB image in this patient who had a middle cerebral artery occlusion confirmed by MRA (24). The lesion is also clearly seen on the ADC, TRACE, eADC, and isoDWI maps; although it is more difficult to observe, slight changes are present on the RA, FA, and VR maps. No change was observed on the T2-w images. These results were typical of our findings in this study. Figure 2Open in figure viewerPowerPoint This individual (patient 5) suffered from a left middle cerebral artery (MCA) territory stroke and was imaged at 5.5 hours from onset. The patient had a baseline NIHSS score of 15 and a 24-hour NIHSS score of 12. This stroke (white arrow in b) was considered a WM stroke. The signal loss seen on some images is attributed to frontal sinus susceptibility loss at 3.0-T (27). Overall, the mean ADC values for normal and infarcted regions were 0.79 × 10–3 mm2 second and 0.45 × 10–3 mm2 second, respectively, demonstrating a –42.4% ± 2.0% change in the ADC during hyper-acute ischemia. When assessing for interactions, only the post-processing strategy was seen to have a significant effect. The effects of infarct location and the interaction term, strategy by infarct location, were not significant. Table 2 and Fig. 3 show the percent difference of each technique averaged across all patients and ROI positions. As expected in hyper-acute stroke, no difference (P > 0.05) was observed in the T2-w percent difference, confirming our decision to use the T2-w image as the comparison standard. Table 2. Percent Difference in the Signal Intensity Between the Ischemic Region and the Contralateral Normal Region* All data (N = 21) White matter (N = 12) Grey matter (N = 4) Both (N = 5) Percent difference P-valuea Percent difference P-valuea Percent differenceb Percent differenceb T2 4.5 ± 2.7 3.6 ± 3.3 2.4 ± 10.6 8.5 ± 3.5 CMB 46.8 ± 4.8 <0.001 43.0 ± 5.2 <0.001 39.3 ± 12.6 61.8 ± 12.0 ADC −42.3 ± 2.0 <0.001 −41.1 ± 2.5 <0.001 −40.1 ± 2.0 −47.6 ± 5.8 TRACE −38.1 ± 2.8 <0.001 −39.7 ± 2.3 <0.001 −28.8 ± 7.9 −41.9 ± 8.0 eADC 40.2 ± 3.8 <0.001 40.0 ± 3.5 <0.001 31.5 ± 9.1 47.4 ± 12.1 isoDWI 50.7 ± 7.9 <0.001 47.8 ± 12.0 <0.001 50.4 ± 12.9 57.8 ± 15.4 RA 0.7 ± 6.5 NS 4.9 ± 10.7 NS −6.8 ± 6.6 −3.1 ± 8.2 FA 20.6 ± 5.5 NS 20.5 ± 9.1 NS 18.6 ± 10.0 22.5 ± 5.4 VR 0.6 ± 7.2 NS −5.2 ± 5.3 NS −2.0 ± 7.0 16.5 ± 28.2 * Statistical significance of the difference was obtained with respect to T2-w image for each technique. Reported percent differences are mean ± standard error. NS = not significant. a P-values represent comparisons to the T2-w image and are adjusted for eight planned comparisons. Using the Bonferroni technique to correct for multiple comparisons, the critical P-value = 0.006. b Because there was no statistical interaction between the various images and the T2-w image, no further analysis was performed, therefore no P-values are available. Figure 3Open in figure viewerPowerPoint Absolute mean percent difference for T2-w images, the five isotropic (CMB, ADC, TRACE, eADC, and isoDWI) post-processing and three anisotropic (RA, FA, and VR) techniques. Strategies are described in Fig. 1. Results are shown for the pooled data as well as broken down by infarct location (WM, GM, and BOTH). Error bars are ± one standard error. All five isotropic post-processing techniques (CMB, ADC, TRACE, eADC, isoDWI) showed statistically significant changes (Fig. 3). The isoDWI technique showed an absolute percent difference of 50.7% ± 7.9%, although the CMB, ADC, TRACE, and eADC techniques all had mean changes > 38%. ADC and TRACE changes were the most significant, and the coefficient-of-variation was the lowest for ADC (0.22) and TRACE (0.33). The RA, FA, and VR techniques did not yield statistically different results. The FA technique had the largest percent change (20.6% ± 5.5%, P > 0.05) of the three anisotropic techniques. While location of infarct was not a statistically significant factor overall, it is of interest to perform an exploratory analysis of the percent difference by location of infarct (Table 2 and Fig. 3). In this analysis, the GM (N = 4) and BOTH (N = 5) groups did not show any interactions between the T2-w image and any of the diffusion images (P > 0.05), so no follow-up, planned comparisons were performed. The WM group (N = 12) showed significant interaction between signal intensity changes and the image construction technique; thus, follow-up, planned comparisons were justified. In the follow-up, planned paired comparisons all of the isotropic images and maps were statistically significant; however, none of the anisotropic techniques achieved statistically. Again, FA had the largest and most consistent changes (20.5% ± 9.1%, P = 0.19). DISCUSSION Various diffusion post-processing strategies have been used in previous studies to assess damage during acute ischemic stroke (2-11, 14). This study compared T2-w images with eight DWI/DTI post-processing strategies. The CMB, ADC, TRACE, eADC, and isoDWI strategies all resulted in statistically significant differences of > 38%. When pooled over all infarct locations, the isoDWI strategy had the largest significant mean difference (50.7%); however, ADC had the highest level of statistical significance and the lowest coefficient of variation (P < 0.001, 0.22, respectively). When analyzed by infarct location, significant changes were found using all five of the isotropic measures (CMB, ADC, TRACE, eADC, and isoDWI) in the predominately WM regions. The analysis by tissue type is exploratory and needs to be treated with caution because the ANOVA model did not show significance by infarct location. The analysis by tissue type has been included since other studies (4, 9) have shown WM anisotropic changes in acute and sub-acute ischemia. Furthermore, the number of patients in each tissue type was small (12 WM strokes, four GM, five BOTH). No interaction was found between the T2-w image and the post-processed images in the GM and BOTH matter strokes. This is most likely due to the small size of these two groups. The larger WM stroke group showed statistical significant changes for only isotropic images and maps. No changes were observed in the T2-w images between infarcted and contralateral normal regions. This was expected during the hyper-acute phase of stroke, as T2 changes are not apparent until after diffusion changes (5). The 42.4% reduction in ADC values between normal and infarcted regions (0.79 × 10–3 mm2 second to 0.45 × 10-3 mm2 second) is consistent with previously published work (c.f., 8,25). This study experimentally confirmed the theoretical equivalence of the ADC and the TRACE maps, because these two maps are essentially the same (mean change of –42.4% ± 2.0% vs.–38.2% ± 2.8%). As with the CMB and isoDWI, the image difference in the ADC and TRACE maps results from differences in noise propagation. The three anisotropic measures (RA, FA, and VR) were not statistically different, although the FA technique showed the largest changes overall and in each of the three sub-group tissue types. The observation in our study that the FA in WM produced a more consistent result is likely due to the fact that it is the least susceptible of the anisotropic techniques to noise propagation (26). The generally irregular performance of the anisotropic techniques differs from that found in other studies of diffusion anisotropy in stroke (4, 6, 7, 9-11). This is likely due to a combination of factors. The first factor is the noise propagation during both the eigen value decomposition (22) process and the relatively complex post-processing steps (described in Fig. 1). Signal averaging has been used in some of the previous studies (4, 7) to improve the results, although this is unlikely to be of use in a rapid (< 15-minute total scan time) acute stroke imaging protocol (14). Second, this study, unlike other studies (4, 6, 7, 9-11), investigated DTI in a hyper-acute stroke population. All of our patients were 1) confirmed as having strokes and 2) imaged between 2.0 and 5.5 hours of onset (mean delay of 3.7 hours ± 1.1 hours). It is possible that more statistically significant changes seen at later time points in other studies may be the result of maturation of the infarct. In the more chronic phases of this disease, a time course of decreasing anisotropy has been reported (10). In the sub-acute and chronic periods, the decrease in anisotropy is likely due to vasogenic edema and the gradual breakdown of cellular membranes (11). In the hyper-acute phase (less than six hours), which we have examined, cytotoxic edema is the dominant mechanism responsible for changes in diffusion (11). Additionally, heterogeneity has been observed in temporal evolution of diffusion anisotropy between subjects and within individual lesions (13). Perhaps a time course for anisotropy exists that peaks between the hyper-acute and acute phase of stroke depending on the severity of the infarct. In the early phase of ischemia the influx of water into the cell lowers the ADC but may not significantly affect diffusion anisotropy. Moreover, Green et al (15) showed that the net direction of diffusion can be altered without significant changes in the FA map. Only the studies by Sorenson et al (4) and Zelaya et al (10) had a mean MR delay of less than 24 hours. Zelaya et al observed a decrease in anisotropy between 12 hours and 90 days after onset. Sorenson et al used a DTI technique with three signal averages on a 1.5-T system. Simplistically, the acquisition differences between Sorenson et al and our method would result in a signal-to-noise ratio that would be approximately 8% to 33% higher (from the averaging) and 30% to 60% lower (from the use of a lower field strength) (27) than the single-average 3.0-T experiments reported here. In the Sorenson et al study, the normalized mean percentage difference in FA between white matter ischemic and normal white matter was 13.5% (P = 0.01) (4), which compares favorably in mean change to our result of 20.5% ± 9.1% (P = 0.19). This study found that in hyper-acute stroke, acquiring the full diffusion tensor and calculating anisotropic maps provides no additional advantage over the isotropic maps available in conventional DWI for detecting ischemia. ADC maps had the most significant percent difference and lowest coefficient of variation of the eight examined strategies. Therefore, in the clinical setting of detecting hyper-acute stroke, there is no apparent benefit in performing a DTI acquisition over a DWI acquisition. Because many hyper-acute stroke patients have difficulty tolerating long scan times and due to therapeutic urgency (14), the shorter DWI technique would be preferred in the development of a rapid stroke imaging protocol. Anisotropic images are designed to investigate the microstructural properties of tissues, so there exists potential in these images to reveal supplemental information regarding the extent or severity of stroke, as has been proposed by Sorenson et al (4) and by Mukherjee et al. (9). This observation may be limited to later time points in ischemic stroke. The strength of anisotropic diffusion images may lie in describing and characterizing the stage of stroke (i.e., cytotoxic vs. vasogenic edema). The absence of anisotropic change may be consistent with early and/or mild ischemia, while increasing anisotropic changes may be indicative of more chronic, more severe stroke. If the time-course of anisotropic changes discussed here are confirmed, anisotropic measures maybe an important method for separating cytotoxic from vasogenic edema and potentially may be of use in helping to define tissue-based therapeutic windows. Because anisotropic maps do not consistently show changes associated with hyper-acute ischemic stroke, isotropic maps obtained using DWI are most appropriate for clinical detection of hyper-acute stroke. Future studies, however, are needed to further examine the changes in DTI during the evolution of stroke. Acknowledgements A.D.H. was an Alberta Heritage Foundation for Medical Research (AHFMR) Summer Student, J.R.M. is a recipient of a Career Development Award from the Multiple Sclerosis Society of Canada and an AHFMR Medical Scholar, M.D.H. is a Heart and Stroke Foundation of Canada (HSFC) Research Scholar, and R.F. is a HSFC Research Scholar and an AHFMR Medical Scholar. The authors thank General Electric Medical Systems for providing the research DTI sequence used in this study; Dr. Jessica Simon and Dr. Carla Wallace for reassessing the stroke location; and Jodi Edwards, MA, for assisting with the statistical analyses. REFERENCES 1 Moseley ME, Cohen Y, Mintorovitch J, et al. Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Reson Med 1990; 14: 330– 346. 2 Schwamm LH, Koroshetz WJ, Sorensen AG, et al. Time course of lesion development in patients with acute stroke, serial diffusion- and hemodynamic-weighted magnetic resonance imaging. Stroke 1998; 29: 2268– 2276. 3 Gonzalez RG, Schaefer PW, Buonanno FS, et al. Diffusion-weighted MR imaging: diagnostic accuracy in patients imaged within 6 hours of stroke symptom onset. Radiology 1999; 210: 155– 162. 4 Sorensen AG, Wu O, Copen WA, et al. Human acute cerebral ischemia: detection of changes in water diffusion anisotropy by using MR imaging. Radiology 1999; 212: 785– 792. 5 Baird AE, Warach S. Magnetic resonance imaging of acute stroke. J Cereb Blood Flow Metab 1998; 18: 583– 609. 6 Higano S, Zhong J, Shrier DA, et al. Diffusion anisotropy of the internal capsule and the corona radiata in association with stroke and tumors as measured by diffusion-weighted MR imaging. AJNR Am J Neuroradiol 2001; 22: 456– 463. 7 Gillard JH, Papadakis NG, Martin K, et al. MR diffusion tensor imaging of white matter tract disruption in stroke at 3 T. Br J Radiol 2001; 74: 642– 647. 8 Desmond PM, Lovell AC, Rawlinson AA, et al. The value of apparent diffusion coefficient maps in early cerebral ischemia. AJNR Am J Neuroradiol 2001; 22: 1260– 1267. 9 Mukherjee P, Bahn MM, McKinstry RC, et al. Differences between gray matter and white matter water diffusion in stroke: diffusion-tensor MR imaging in 12 patients. Radiology 2000; 215: 211– 220. 10 Zelaya F, Flood N, Chalk JB, et al. An evaluation of the time dependence of the anisotropy of the water diffusion tensor in acute human ischemia. Magn Reson Imaging 1999; 17: 331– 348. 11 Yang Q, Tress BM, Barber PA, et al. Serial study of apparent diffusion coefficient and anisotropy in patients with acute stroke. Stroke 1999; 30: 2382– 2390. 12 Frayne R, Sevick RJ, Demchuk AM, et al. Clinical stroke imaging at 3 T. In: Proceedings of the 8th Annual Meeting of ISMRM, Denver, 2000. p 1253. 13 Sotak CH. The role of diffusion tensor imaging in the evaluation of ischemic brain injury—a review. NMR Biomed 2002: 15: 561– 569. 14 Sunshine JL, Tarr RW, Lanzieri CF, Landis DM, Selman WR, Lewin JS. Hyperacute stroke: ultrafast MR imaging to triage patients prior to therapy. Radiology 1999; 212: 325– 332. 15 Green HAL, Pena A, Price CJ, Warburton EA, Pickard JD, Carpenter TA, Gillard JH. Increased anisotropy in acute stroke, a possible explanation. Stroke 2002; 33: 1517– 1521. 16 Papadakis NG, Xing D, Huang CLH, Hall LD, Carpenter TA. A comparative study of acquisition schemes for diffusion tensor imaging using MRI. J Magn Reson 1999; 137: 67– 82. 17 Barber PA, Demchuk AM, Zhang J, Buchan AM. Validity and reliability of a quantitative computed tomography score in predicting outcome of hyperacute stroke before thrombolytic therapy. Lancet 2000; 355: 1670– 1674. 18 Hill MD, Barber PA, Demchuk AM, et al. Building a “brain attack” team to administer thrombolytic therapy for acute ischemic stroke. CMAJ 2000; 162: 1589– 1593. 19 Powers DW. Assessment of the stroke patient using the NIH stroke scale. Emerg Med Serv 2001; 30: 52– 56. 20 Conturo TE, McKinstry RC, Akbudak E, Robinson BH. Encoding of anisotropic diffusion with tetrahedral gradients: a general mathematical diffusion formalism and experimental results. Magn Reson Med 1996; 35: 399– 412. 21 Basser PJ, Shrager R. Anisotropically weighted MRI. In: Proceedings of the 5th Annual Meeting of ISMRM, Vancouver, Canada, 1997. p 226. 22 Basser PJ, Pierpaoli C. A simplified method to measure the diffusion tensor from seven MR images. Magn Reson Med 1998; 39: 928– 934. 23 Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn Reson 1996; 111B: 209– 219. 24 Yang JJ, Hill MD, Morrish WF, et al. Comparison of pre- and post-contrast 3D TOF MR angiography for distal intra-cranial branch occlusions in acute ischemic stroke. AJNR Am J Neuroradiol 2002; 23: 557– 567. 25 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. 26 Papadakis NG, Xing D, Houston GC, et al. A study of rotationally invariant and symmetric indices of diffusion anisotropy. Magn Reson Imaging 1999; 17: 881– 892. 27 Frayne R, Goodyear BG, Lauzon ML, Sevick RJ. Magnetic resonance imaging at 3.0 T: technical challenges and benefits in neurological imaging. Invest Radiol 2003; 38: 436– 442. Citing Literature Volume20, Issue2August 2004Pages 193-200 FiguresReferencesRelatedInformation" @default.
• W2100064617 created "2016-06-24" @default.
• W2100064617 creator A5054532501 @default.
• W2100064617 creator A5064182609 @default.
• W2100064617 creator A5073479052 @default.
• W2100064617 creator A5084100416 @default.
• W2100064617 creator A5085398611 @default.
• W2100064617 creator A5091805367 @default.
• W2100064617 date "2004-01-01" @default.
• W2100064617 modified "2023-09-26" @default.
• W2100064617 title "A comparison of images generated from diffusion-weighted and diffusion-tensor imaging data in hyper-acute stroke" @default.
• W2100064617 cites W1967022145 @default.
• W2100064617 cites W1981340850 @default.
• W2100064617 cites W2001037511 @default.
• W2100064617 cites W2032011764 @default.
• W2100064617 cites W2035023461 @default.
• W2100064617 cites W2042462361 @default.
• W2100064617 cites W2043151182 @default.
• W2100064617 cites W2049965228 @default.
• W2100064617 cites W2060327991 @default.
• W2100064617 cites W2067739098 @default.
• W2100064617 cites W2071778792 @default.
• W2100064617 cites W2099708759 @default.
• W2100064617 cites W2103048292 @default.
• W2100064617 cites W2104608904 @default.
• W2100064617 cites W2131096301 @default.
• W2100064617 cites W2138394070 @default.
• W2100064617 cites W2139158372 @default.
• W2100064617 cites W2148656951 @default.
• W2100064617 cites W2150057120 @default.
• W2100064617 doi "https://doi.org/10.1002/jmri.20116" @default.
• W2100064617 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/15269943" @default.
• W2100064617 hasPublicationYear "2004" @default.
• W2100064617 type Work @default.
• W2100064617 sameAs 2100064617 @default.
• W2100064617 citedByCount "37" @default.
• W2100064617 countsByYear W21000646172012 @default.
• W2100064617 countsByYear W21000646172013 @default.
• W2100064617 countsByYear W21000646172014 @default.
• W2100064617 countsByYear W21000646172017 @default.
• W2100064617 countsByYear W21000646172018 @default.
• W2100064617 countsByYear W21000646172022 @default.
• W2100064617 countsByYear W21000646172023 @default.
• W2100064617 crossrefType "journal-article" @default.
• W2100064617 hasAuthorship W2100064617A5054532501 @default.
• W2100064617 hasAuthorship W2100064617A5064182609 @default.
• W2100064617 hasAuthorship W2100064617A5073479052 @default.
• W2100064617 hasAuthorship W2100064617A5084100416 @default.
• W2100064617 hasAuthorship W2100064617A5085398611 @default.
• W2100064617 hasAuthorship W2100064617A5091805367 @default.
• W2100064617 hasBestOaLocation W21000646171 @default.
• W2100064617 hasConcept C121332964 @default.
• W2100064617 hasConcept C126322002 @default.
• W2100064617 hasConcept C126838900 @default.
• W2100064617 hasConcept C143409427 @default.
• W2100064617 hasConcept C149550507 @default.
• W2100064617 hasConcept C153180895 @default.
• W2100064617 hasConcept C154945302 @default.
• W2100064617 hasConcept C2776572282 @default.
• W2100064617 hasConcept C2780645631 @default.
• W2100064617 hasConcept C3020166492 @default.
• W2100064617 hasConcept C41008148 @default.
• W2100064617 hasConcept C69357855 @default.
• W2100064617 hasConcept C71924100 @default.
• W2100064617 hasConcept C97355855 @default.
• W2100064617 hasConceptScore W2100064617C121332964 @default.
• W2100064617 hasConceptScore W2100064617C126322002 @default.
• W2100064617 hasConceptScore W2100064617C126838900 @default.
• W2100064617 hasConceptScore W2100064617C143409427 @default.
• W2100064617 hasConceptScore W2100064617C149550507 @default.
• W2100064617 hasConceptScore W2100064617C153180895 @default.
• W2100064617 hasConceptScore W2100064617C154945302 @default.
• W2100064617 hasConceptScore W2100064617C2776572282 @default.
• W2100064617 hasConceptScore W2100064617C2780645631 @default.
• W2100064617 hasConceptScore W2100064617C3020166492 @default.
• W2100064617 hasConceptScore W2100064617C41008148 @default.
• W2100064617 hasConceptScore W2100064617C69357855 @default.
• W2100064617 hasConceptScore W2100064617C71924100 @default.
• W2100064617 hasConceptScore W2100064617C97355855 @default.
• W2100064617 hasIssue "2" @default.
• W2100064617 hasLocation W21000646171 @default.
• W2100064617 hasLocation W21000646172 @default.
• W2100064617 hasOpenAccess W2100064617 @default.
• W2100064617 hasPrimaryLocation W21000646171 @default.
• W2100064617 hasRelatedWork W1995041400 @default.
• W2100064617 hasRelatedWork W2058081075 @default.
• W2100064617 hasRelatedWork W2082520993 @default.
• W2100064617 hasRelatedWork W2251032337 @default.
• W2100064617 hasRelatedWork W2288149305 @default.
• W2100064617 hasRelatedWork W2330171289 @default.
• W2100064617 hasRelatedWork W2355091906 @default.
• W2100064617 hasRelatedWork W2611590172 @default.
• W2100064617 hasRelatedWork W2936043023 @default.
• W2100064617 hasRelatedWork W2425903266 @default.
• W2100064617 hasVolume "20" @default.
• W2100064617 isParatext "false" @default.
• W2100064617 isRetracted "false" @default.
• W2100064617 magId "2100064617" @default.

• next