This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yang, Q. Articles by Davis, S. M. Search for Related Content PubMed PubMed Citation Articles by Yang, Q. Articles by Davis, S. M.

(Stroke. 1999;30:2382-2390.)
© 1999 American Heart Association, Inc.

Original Contributions

Serial Study of Apparent Diffusion Coefficient and Anisotropy in Patients With Acute Stroke

Presented in part at the Annual Meeting of the American Society of Neuroradiology, San Diego, Calif, May 23–28, 1999.

Qing Yang, PhD; Brian M. Tress, MD, FRACR; P. Alan Barber, FRACP; Patricia M. Desmond, MSc, FRACR; David G. Darby, PhD, FRACP; Richard P. Gerraty, MD, FRACP; Ting Li, PhD Stephen M. Davis, MD, FRACP

From the Departments of Radiology (Q.Y., B.M.T., P.M.D, T.L.) and Neurology (P.A.B., D.G.D., R.P.G., S.M.D), Royal Melbourne Hospital and University of Melbourne, Victoria, Australia.

Correspondence to Dr Qing Yang, Department of Radiology, University of Melbourne, Royal Melbourne Hospital, Parkville, Victoria 3050, Australia. E-mail QY{at}radior.medrmh.unimelb.edu.au

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Background and Purpose—We sought to characterize the evolution of apparent diffusion coefficient (ADC) and apparent diffusion anisotropy (ADA) in acute stroke and to evaluate their roles in predicting stroke evolution and outcome.

Methods—We studied 26 stroke patients acutely (<24 hours), subacutely (3 to 5 days), and at outcome (3 months). Ratios of the ADC and ADA within a region of infarction and the normal contralateral region were evaluated and compared with the Canadian Neurological Scale, Barthel Index, and Rankin Scale.

Results—Heterogeneity in ADC and ADA evolution was observed not only between patients but also within individual lesions. Three patterns of ADA evolution were observed: (1) elevated ADA acutely and subacutely; (2) elevated ADA acutely and reduced ADA subacutely; and (3) reduced ADA acutely and subacutely. At outcome, reduced ADA with elevated ADC was observed generally. We identified 3 phases of diffusion abnormalities: (1) reduced ADC and elevated ADA; (2) reduced ADC and reduced ADA; and (3) elevated ADC and reduced ADA. The ADA ratios within 12 hours correlated with the acute Canadian Neurological Scale (r=0.46, P=0.06), subacute Canadian Neurological Scale (r=0.55, P=0.02), outcome Barthel Index (r=0.62, P=0.01), and Rankin Scale (r=-0.77, P<0.0005) scores.

Conclusions—Combined ADC and ADA provide differential patterns of stroke evolution. Early ADA changes reflect cellular alterations in acute ischemia and may provide a potential marker to predict stroke outcome.

Key Words: cerebral edema • cerebral infarction • magnetic resonance imaging, diffusion-weighted • stroke, ischemic • stroke outcome

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Diffusion-weighted imaging (DWI) has become an important MRI method for the early diagnosis and characterization of ischemic stroke.^{1 2 3 4 5 6 7 8 9 10} Early reports^{1 2} showed that acute infarcts appear hyperintense on DWI because of a reduction in the apparent diffusion coefficient (ADC) of water within minutes after the onset of ischemia,³ reflecting early disruption of energy metabolism.

However, the mechanism behind the ADC reduction in acute ischemia is not fully understood. In biological tissue, the ADC of water molecules is much lower than its free water value because of physical restrictions from membranes, fibers, and macromolecules such as proteins.¹¹ The amount of dissolved organic molecules may also alter the viscosity of water. In acute ischemia, disruption of energy metabolism with failure of ion pumps causes cell swelling and accumulation of intracellular sodium and water (cytotoxic edema).^{7 8} The net migration of water from extracellular space (ECS, where ADC is presumed to be high) into intracellular space (ICS, where ADC is presumed to be low) has been considered the dominant mechanism for the ADC reduction.^{1 6} Another postulated mechanism is that the ADC reduction is a result of decreased membrane permeability caused by collapse of transmembrane ion gradients.¹² Furthermore, recent studies have provided compelling evidence that the observed ADC reduction is dominated by the water ADC reduction in the ICS because of a decrease in energy-dependent cytoplasmic circulation or an increase in water viscosity.^{13 14 15} On the other hand, the time course correlation of the ADC reduction with the decrease in ECS volume and the increase in ECS tortuosity suggests that cell swelling may also cause ADC reduction of water or other molecules in the ECS.^{15 16} Recent numerical modeling has suggested both cellular swelling and membrane permeability to be important factors influencing ADC in spinal cord white matter.¹⁷

The temporal evolution of diffusion abnormalities is also important in tracking stroke progression. Studies using animal models have revealed profound ADC changes at early acute stages.^{2 3 4} Moreover, there is an increase in the lesion volume over time as measured by decreased ADC values.³ In human stroke, while some studies^{6 7} report the persistence of a reduced ADC for at least 4 days after stroke onset, others have found heterogeneity of ADC within the infarct while part or all of the lesion displayed pseudonormal or high ADC values by 24 to 48 hours.^{5 18} This disparity may be related to different patient populations or stroke etiologies,^{5 18} different data acquisition and analysis techniques,¹⁹ and anisotropic diffusion effects.^{20 21} The diffusion of water molecules in biological tissue is anisotropic, particularly in tissue containing nerve fibers and white matter tracts, where ADC is high in the direction parallel to fiber tracts and low in perpendicular directions.^{22 23} Anisotropic diffusion can cause conspicuous hyperintensities on DWI that may be confused with acute ischemia⁹ and influence the accurate quantitation of ADC.^{20 21} Such effects can be minimized by calculating the trace ADC and isotropic DWI images from DWI measurements in 3 orthogonal directions.^{20 24} Recent studies^{25 26 27} have evaluated the time course of the trace ADC in human stroke, eliminating the influences of anisotropic diffusion. Computer-assisted segmentation analysis has further demonstrated ADC heterogeneity within the infarct and partial ADC elevation by 5 to 9 hours after ictus, although the average ADC in the lesion remained low.²⁷

Measurements of apparent diffusion anisotropy (ADA) may provide additional information about cellular changes in ischemic stroke.^{22 23 28 29 30 31 32 33} To our knowledge, no systematic study of the evolution of combined ADC and ADA in acute human stroke has been reported. In this study we evaluated the serial changes of ADC and ADA abnormalities in patients with acute ischemic stroke. We sought to improve our understanding of the mechanisms underlying the ADC and ADA changes in ischemia and to determine whether these data are useful in characterizing and predicting the evolution of acute ischemic stroke.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Patients with sudden onset of focal neurological deficit consistent with hemispheric ischemic stroke were recruited from the Stroke Unit of the Royal Melbourne Hospital. Stroke onset was defined as the last time the patient was known to be without neurological deficit. Patients were excluded if they had cerebral hemorrhage, preexisting significant nonischemic neurological deficit (including dementia or extrapyramidal disease), or a history of prior stroke that would hinder interpretation of clinical and radiological data. Patients treated by putative neuroprotective or thrombolytic drugs were excluded. The study was performed with the approval of the ethics committee at our institution, and written informed consent was obtained from the patient or next of kin.

Clinical Assessment
The Canadian Neurological Scale (CNS), a validated neurological impairment score,³⁴ was measured just before the acute and subacute MRI studies. Outcome clinical assessments were performed on the same day as the outcome MRI study and consisted of a repeated CNS and scores derived from the Barthel Index (BI) and the Rankin Scale (RS).³⁵ The BI is a validated functional disability score, and the RS is a validated handicap scale. These clinical scales were used because they measure different aspects of recovery after stroke. All clinical assessments were performed by a neurologist or neurology resident trained in their administration and were administered without knowledge of the MRI results.

Imaging Parameters
All patients were scanned on a 1.5-T clinical whole-body scanner (Signa Horizon SR120, GE Medical Systems) with the use of an optimized protocol including a T1-weighted sagittal localizer, DWI sequence, contrast-enhanced perfusion imaging (CEPI) sequence, dual proton-density and T2-weighted fast-spin-echo sequence, spin-echo echo planar imaging (EPI) sequence, phase-contrast MR angiography (MRA), and finally a contrast-enhanced T1-weighted sequence. Similar slice locations were used to facilitate comparisons. The total "table time" for these sequences was approximately 20 minutes. Only the DWI results are reported in this study.

DWI scans were performed with a single-shot, spin-echo EPI sequence with the Stejskal-Tanner diffusion-encoding method.³⁶ The DWI parameters were as follows: 40x20-cm field of view, 256x128 matrix size, 16 axial slices, 6-mm slice thickness, and 1-mm gap covering the whole brain. The first 19 patients were studied with a trace DWI sequence with 5 diffusion b values (0 to 1000 s/mm²) in each of 3 orthogonal directions²⁴ and repetition time/echo time (TR/TE) of 6000/110 ms. The remaining 7 patients had diffusion tensor imaging (DTI)^{22 23} with 3 b values (0 to 1000 s/mm²) in each of 6 directions and TR/TE of 10 000/110 ms. Scanning time was 1 minute 18 seconds for the trace DWI and 2 minutes 10 seconds for the DTI.

Image Processing
Postprocessing of images was performed on a UNIX workstation with the use of customized software developed in IDL (Interactive Data Language, Research Systems Inc). Since the calculation of diffusion anisotropy is sensitive to image noise and artifacts, noise reduction and correction for image distortions were performed on all images. Raw images were filtered with a 9x9 gaussian kernel (=0.5). The average background noise was subtracted to reduce the nonlinear influences on the DWI signal attenuation. The EPI sequence has first-order eddy current compensation to minimize image distortions. However, higher-order eddy current distortions due to the strong diffusion gradients may still cause artifacts, particularly on ADA quantification. Therefore, corrections for such distortions were applied with a modified approach based on the translation-shear-scaling model.³⁷ Although the single-shot EPI sequence eliminated motion artifacts from each "snapshot" image, possible head movement during the DWI scanning time (1 to 2 minutes) would be captured, causing mismatch between images. Such mismatch may vary between images and slices and cannot be easily corrected with the use of standard rigid-body coregistration algorithms. Therefore, any data set with interimage mismatch was aligned by a dynamic visual-manual adjustment based on both ratio and difference maps, followed by an automated fine-tuning approach. All image corrections employed a cubic convolution interpolation method, which closely approximates the theoretically optimum sinc interpolation function³⁸ to preserve image resolution. Motion and distortion artifacts were double-checked in animation mode between images of different b values in each direction and between images of the same b value at different directions. The DWI signal intensity attenuation curve was also dynamically checked, particularly in the region of interest (ROI). Any individual image with noticeable artifacts was excluded from the fitting process for calculation of the ADC map in each of the 3 (for trace DWI) or 6 directions (for DTI).

Both the trace DWI and the DTI allow the calculation of ADCs in each of 3 orthogonal directions, which then provide the average ADC (ADC_av, noted as ADC hereafter) and the orientation-dependent standard deviation index of ADA (ADA_sd).²⁴ In addition, DTI allows the calculation of the full diffusion tensor²² and hence a more accurate orientation-independent anisotropy index, such as the fractional anisotropy (ADA_fr).²³ The trace DWI and DTI sequences were tested on a standard water phantom at room temperature. The calculated ADC value of free water was 2.2±0.1 x10^-3 mm²/s, with an ADA_sd value of 0.04±0.02 and ADA_fr value of 0.16±0.05. These non-zero ADA values are mainly due to signal-to-noise ratio.²² Both ADA indices range from 0 to 1, representing completely isotropic to extreme anisotropic diffusion.

Data Analysis
Heterogeneity of ADC within the lesion has been reported⁵ and further confirmed with computer-assisted segmentation analysis.²⁷ Our experience confirmed this finding. The heterogeneous ADC and ADA distribution within the lesion can also be visualized by adjusting the image contrast level and window, particularly within large infarctions. Therefore, segmentation analysis of the lesion is necessary to properly study the evolution of ADC and ADA in different tissue areas. Automated segmentation analysis requires specialized computer software,^{5 27} which is not widely available. Furthermore, it would be difficult to automatically follow the same tissue area over serial studies because of the heterogeneous progression and edematous swelling of the lesion. Therefore, we used a segmentation approach based on anatomic location and tissue type similar to that used in animal model studies.^{2 3} Since different tissue structures have different ADA values, proper evaluation of ADA changes needs to be linked to specific tissue types. The ADA of the white matter is much greater than that of the gray matter, and the ADA map is very useful in delineating white matter. When one considers the image resolution and slice thickness of the DWI images, partial volume effects (PVE) were inevitable, particularly between the gray matter and the white matter or the cerebrospinal fluid (CSF) in cerebral gyral areas. Therefore, lesions were segmented and grouped as white matter (WM), cortical or deep gray matter (GM), and mixed gyral gray/white matter (GWM) regions. The isotropic DWI (with b=1000), T2-weighted image (which is DWI with b=0), and ADC and ADA maps of serial studies were visualized simultaneously. A ROI was initially selected by free-hand tracing on the acute ADC map to outline the infarct with reduced ADC. The ROI was then projected on other images and further edited for reliable segmentation, guided by anatomic knowledge and histogram analysis. No serial measurement was performed for later expanded areas of infarction. Efforts were made to track each segmented tissue area with all 4 kinds of images, guided by anatomic knowledge, while edematous swelling was also considered. In a few cases in which asymmetric head angulation caused difficulties for comparison with the contralateral side on the same slice, multiple slices were analyzed. Additionally, it was not practical to reproduce the exact same slice locations over serial MRI studies for all patients, especially when head movements occurred during the scan, although a standard landmark was always used. Thus, some raw DWI data may be subject to PVE within half the slice thickness. However, efforts were made to minimize additional PVE during the data analysis. No segmentation was performed for small infarctions, which were evaluated separately from large lesions. The average ADC and ADA values within each segmented ROI were obtained with histogram analysis to eliminate contamination from individual noise. Care was also taken to avoid contamination from sulcal CSF. To attain a reliable measure of the evolution of the ADC and ADA changes between patients and over serial studies, ADC and ADA ratios were calculated by dividing the mean values in the lesion ROI by those in the corresponding contralateral ROI.²⁰

Linear regression analyses were performed between the ADC and ADA ratios and the clinical scores (CNS, BI, RS) with Pearson's correlation coefficient (r) and significance level (P) of the F test. Results were considered statistically significant at levels of P<0.05. Since the evolutions of ADC and ADA are heterogeneous and complicated, no statistical analysis was performed to test for ratio difference, but all data points are presented.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Twenty-six patients (16 men, 10 women; mean age, 69±12 years; range, 42 to 92 years) were studied, with 9 small lesions (lacunar, striatocapsular, and small cortical infarctions) and 17 large lesions (major arterial territory infarctions). All 26 patients had acute studies, and 25 had subacute studies. One patient missed the subacute study but completed the outcome study. Twenty patients (including all 9 patients with small lesions) had outcome studies. Three patients refused, and 1 was unable to tolerate the outcome MRI scan, but all of them had the outcome clinical assessments. One patient died as a result of complications from stroke (assigned outcome BI=5 and RS=0), and 1 died of an unrelated cardiac event after the subacute study (outcome BI and RS not available). Time from stroke onset to the acute study was 11.5±7.2 hours (range, 2.5 to 23.5 hours), with 10 studies within 6 hours and 17 studies within 12 hours. Time to the subacute study was 3.7±1.2 days (range, 1.9 to 6.8 days). Time to the outcome study was 90±26 days (range, 35 to 154 days).

Trace DWI and DTI
A total of 71 MRI studies were performed with 51 trace DWI and 20 DTI scans. The mean ADC value in all normal ROIs was 0.85±0.14 (x10^-3 mm²/s), which agrees well with other reports.^{20 25 26 27}

All 71 DWI scans provided ADA_sd maps, while 20 DTI scans provided additional ADA_fr maps. Since ADA_sd is orientation dependent, it underestimates the diffusion anisotropy of water in tissues depending on tissue types and fiber orientations relative to the 3 orthogonal diffusion-encoding directions.²² However, our experience indicated similar ADA_sd and ADA_fr changes in terms of elevation or reduction by comparing the ischemic lesion with the corresponding contralateral region. To further clarify this phenomenon, the mean values of both indices in the segmented infarcts and normal ROIs (including CSF) from the 20 DTI scans were plotted, as shown in Figure 1a. This illustrates the range of quantitative ADA values of different tissue types measured by both indices, in good agreement with other reports.^{22 30 33} The non-zero ADA values in CSF agree well with those measured in a water phantom. Figure 1a demonstrates the empirical relationship between ADA_sd and ADA_fr, as described by a fitted curve: ADA_fr=1-exp(-4.2 ADA_sd). A 2-dimensional histogram analysis of the associated ADA_sd and ADA_fr maps further supported such an approximate relationship. The ADA_sd values were converted into ADA_sd* values, which are approximately equivalent to ADA_fr according to the above relationship. The ratios of ADA_sd* in the lesion to that in the contralateral region were calculated and compared with the corresponding ADA_fr ratios, as shown in Figure 1b. A 1:1 relationship between the ADA_fr and the corrected ADA_sd* ratios was found by linear regression analysis (slope=1.00±0.02, r=0.94, P<10^-12). This allowed us to extend the above correction to all the ADA_sd values measured by the trace DWI. The corrected ADA_sd* ratios were included together with the ADA_fr ratios measured by the DTI to increase statistical power in subsequent analysis.

View larger version (16K): [in this window] [in a new window]   Figure 1. Relationship between ADAfr and ADAsd values (a) and between ADAfr and ADAsd* ratios (b), as described in the text, for WM (rectangle), GM (circle), GWM (diamond), and CSF (star). Solid and open symbols represent normal and infarcted tissues, respectively. The solid curve in panel a represents the empirical relationship ADAfr=1-exp(-4.2 ADAsd). The solid line in panel b shows a 1:1 relationship supported by linear regression analysis.

ADC and ADA Evolution in Small Infarctions
The time courses of the ADC and ADA ratios in 9 patients with small lesions are shown in Figure 2. All 9 acute lesions displayed reduced ADC; 8 had elevated ADA, while 1 WM lesion at 23.5 hours had low ADA ratio. Subacutely, the ADC ratios tended toward normal in 2 small cortical GM lesions and remained low in the others. The initially elevated ADA ratios persisted in 2 WM lesions, became reduced in 4 WM lesions, and approached normal in the 3 small cortical GM lesions. At outcome, the ADC ratio became elevated in 8 patients, while 1 had persistent ADC reduction at 85 days. The ADA ratio became further reduced in all lesions. Overall, the ADC and ADC changes during infarct evolution were greater in WM than in GM lesions.

View larger version (26K): [in this window] [in a new window]   Figure 2. Time course of ADC and ADA ratios in small lesions after stroke onset. Solid lines join the serial data points for each patient. Time is on logarithmic scale.

ADC and ADA Evolution in Large Infarctions
A total of 69 segmented lesion ROIs were sampled in serial studies of 17 patients. The time courses of the ADC and ADA ratios in all the segmented ROIs are shown in Figure 3. General patterns of ADC and ADA evolution are similar to those in the small lesions of the same tissue types, except that no acutely elevated ADA was observed in the deep GM. The evolution of ADC and ADA ratios in the GWM tissues is also similar to that of small WM lesions. However, earlier ADC pseudonormalization or elevation by 24 to 48 hours can be seen in both WM and GM lesions. Figure 4 demonstrates the serial ADC and ADA maps of a patient with initially reduced ADC and elevated ADA in the WM-dominated lesion. These evolved to elevated ADC and reduced ADA by 42 hours after stroke onset. Heterogeneous ADC distribution within the lesion can be seen. In addition, the initial lesion expanded by 42 hours, with a peripheral rim of reduced ADC and elevated ADA resembling the initial diffusion abnormalities of the infarct core. At 93 days, both the infarct core and the rim displayed different levels of ADC elevation and ADA reduction (Figure 4). Such expanded infarct areas were not included in the serial analysis. Figure 5 demonstrates differences in both ADC and ADA evolutions within the same lesion, in which the initial infarct had relatively homogeneous ADC reduction and ADA elevation but displayed significant heterogeneities between the WM and peripheral GWM areas at 5 days. Although there are some differences between the ADA_fr and ADA_sd maps, similar features of ADA changes in the lesion can be seen compared with the contralateral side. This patient's head was tilted to the left at the subacute stage, reflecting the actual clinical situation. However, no severe head movement occurred during the DWI sequence, and any interimage mismatch was eliminated.

View larger version (37K): [in this window] [in a new window]   Figure 3. Time course of ADC and ADA ratios in segmented large lesions after stroke onset. Solid lines join the serial data points for each segmented ROI. The solid rectangle and diamond correspond to the segmented WM and GWM lesions in the patient who had slow ADC and ADA recovery, as shown in Figure 6.

View larger version (147K): [in this window] [in a new window]   Figure 4. Serial ADC (top row) and ADAsd (bottom row) maps of a patient who displayed an early ADC elevation of the infarct core with later development of ischemia in the peripheral rim by 42 hours. Note heterogeneity of ADC within the lesion and different ADC levels between the core and periphery at outcome.

View larger version (156K): [in this window] [in a new window]   Figure 5. Serial ADC (top row), ADAsd (middle row), and ADAfr (bottom row) maps of a patient demonstrating heterogeneity in ADC and ADA evolution of the ischemic lesion. The initial infarct had relatively homogeneous ADC reduction and ADA elevation but displayed significant heterogeneities between the WM and the peripheral GWM areas at 5 days. The ADAsd and ADAfr maps display similar features of ADA changes. Note that the tilted head position reflects actual clinical situation; however, motion artifacts have been eliminated (see imaging processing section).

For most of the small and segmented large lesions, 3 patterns of ADA evolution were observed generally: (1) elevated ADA acutely and subacutely; (2) elevated ADA acutely and reduced ADA subacutely; and (3) reduced ADA acutely and subacutely. At outcome, elevated ADC with reduced ADA was observed generally. The exceptional cases, including 1 WM lesion (Figure 3, solid squares in top row) and 3 small cortical GM lesions (Figure 2, bottom row), had both ADC and ADA ratios slightly different from normal by outcome studies. Figure 6 demonstrates the serial ADC and ADA maps, isotropic DWI, and T2-weighted image in the WM lesion case. Despite the usual progression of the peripheral GWM lesion area (white arrowhead in Figure 6, solid diamonds in middle row of Figure 3), the acute WM lesion (white arrow) progressed gradually, approaching normal ADC and ADA and slight T2-weighted image hyperintensity (black arrow) by outcome. This is the only case in this study in which the control ADC and ADA values were sampled from surrounding nonlesional tissues to avoid an old lesion on the contralateral side (double black arrows), which might have led to less reliable ratios. The absolute ADC values were 0.56, 0.73, and 0.87 (x10^-3 mm²/s) in the WM lesion and 0.71, 0.78, and 1.21 (x10^-3 mm²/s) in the peripheral GWM lesion area at 4.5 hours, 51 hours, and 85 days, respectively. The absolute ADC value was 0.86±0.05 (x10^-3 mm²/s) in the normal tissues of this patient, which confirmed the slow ADC recovery in the WM lesion area. Examination of the MRA and CEPI results in this patient indicated early reperfusion. However, the slight T2 hyperintensity in this WM area at outcome suggests some permanent pathological changes.

View larger version (109K): [in this window] [in a new window]   Figure 6. Serial ADC (top row) and ADAfr (second row) maps, isotropic DWI (third row), and T2-weighted image (bottom row) images of a patient demonstrating slow ADC recovery of the acute WM lesion (arrow on acute DWI), while the peripheral GWM lesion (arrowhead) progressed to elevated ADC by outcome. Note slight hyperintensity on the outcome T2-weighted image (black arrow) and an old lesion on the contralateral side (double black arrows).

Relationship Between ADC and ADA
The combined ADC and ADA ratios in all lesions over serial studies are shown in Figure 7. Three phases of diffusion abnormalities are distinguishable: (1) elevated ADA and reduced ADC; (2) reduced ADA and reduced ADC; and (3) reduced ADA and elevated ADC. Most infarctions displayed a transition from phases 1 to phase 2 between the acute and subacute stages, then to phase 3 by outcome. However, phase 3 had occurred by the subacute stage in some WM and GM regions of large lesions, while phase 2 persisted from the subacute to outcome stages in 1 small WM lesion. It is noted that the 3 small cortical GM lesions (solid circle in Figure 7) and the slowly evolving WM lesion (Figure 6, solid square in Figure 7) tended to transform from phase 1 via the normal ADC and ADA cross point into phase 3. For all small and large lesions, there is a correlation between ADC and ADA ratios for transition from phase 1 to phase 2 (r=-0.41, P=0.0005) and from phase 2 to phase 3 (r=-0.52, P<0.00005). However, the overall transition between all 3 phases is nonlinear, presumably reflecting different pathological processes during infarct evolution.

View larger version (18K): [in this window] [in a new window]   Figure 7. Plot of ADA vs ADC ratios for all WM (square), GM (circle), and GWM (diamond) lesions over serial measurements. Solid symbols relate to special cases discussed in the text. Solid lines are linear regression results for transitions from phase 1 to 2 (r=-0.41, P=0.0005) and from phase 2 to 3 (r=-0.52, P<0.00005), respectively.

Correlation With Clinical Score
For large infarctions with multiple segmented ROIs, the average ADC and ADA ratios weighted by the segmented tissue volume were calculated to represent an average diffusion abnormality of the entire lesion slice. Correlation analyses were performed between the patient's clinical scores (CNS, BI, RS) and the average ADC and ADA ratios for the time intervals of <12 hours, 12 to 24 hours, 2 to 10 days (subacute stage), and >35 days (outcome). Correlations were found between ADA ratios within 12 hours of stroke onset and the acute CNS (r=0.46, P=0.06), subacute CNS (r=0.55, P=0.02), outcome BI (r=0.62, P=0.01), and RS (r=-0.77, P<0.0005) scores. Additional analysis confirmed there was no significant correlation between the ADA ratios and the entire acute DWI lesion volume or outcome T2 lesion volume. This suggests that the correlations between the ADA ratio and clinical scores are not biased by the lesion size. Furthermore, there was no significant correlation between clinical scores and ADA ratios at later stages or ADC ratios at any stage. This suggests that early ADA changes may provide an important marker in predicting stroke outcome.

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
In this study of patients with acute ischemic stroke, we have found different patterns of ADC and ADA evolution over time and different phases of combined ADC and ADA abnormalities. In addition, we have observed differences in both ADC and ADA evolution not only between patients but also within individual lesions. Furthermore, we have found that ADA ratios within 12 hours of stroke onset correlate significantly with acute, subacute, and outcome clinical scores.

The combined ADC and ADA information may provide insight into cellular changes during ischemia evolution. In acute ischemia, early disruption of energy metabolism leads to failure of transmembrane ion pumps and cell swelling (cytotoxic edema).³⁹ As infarction evolves, vasogenic edema may develop as a result of the blood-brain barrier breakdown, causing excessive water accumulation and tissue swelling.⁴⁰ However, both cytotoxic and vasogenic edema may present simultaneously in ischemic lesions, and early vasogenic edema may not always cause ECS swelling.⁴ Thus, it is helpful to regard the progression of an ischemic lesion with specific compartment-related swelling.

Different ADC models could be applied in attempt to explain the combined ADC and ADA changes. In the ECS model, the ADC of water is presumed to be high in the ECS and low in the ICS, with the ADC reduction in acute ischemia explained by the net shift of water from the ECS into the ICS, as a result of cell swelling in cytotoxic edema.^{1 6 24} In addition, the shrinkage of the ECS with increased ECS tortuosity²⁸ may cause increased restriction of extracellular water movement, leading to elevated ADA.³¹ While this may qualitatively explain the reduced ADC with elevated ADA in phase 1, excessive water accumulation and ECS swelling in vasogenic edema would lead to reduced ADA accompanied by elevated ADC (phase 3). However, reduced ADC and reduced ADA (phase 2) were observed in many cases with T2-weighted image hyperintensities and extensive tissue distortions, suggesting excessive water accumulation and ECS swelling due to vasogenic edema. Thus, phase 2 could not be easily explained by this model.

In other models, the ADC reduction in acute ischemia may be dominated by decreased ADC in the ICS due to decreased membrane permeability¹² or cytoplasmic circulation or increased viscosity of water.^{13 14 15} In these ICS models, the ADC of water in the ICS and ECS may be similar.¹⁵ To explain our results of combined ADC and ADA changes, we postulate that elevated ADA may reflect enhanced restriction of intracellular water movement, due to decreased membrane permeability, or enhanced ICS weighting, due to water shift from the ECS into ICS caused by cell swelling. However, cell swelling may cause reduced restriction of ICS water movement and hence a reduction of the elevated ADA. The level of ADA elevation may reflect the degree of cellular swelling and membrane degradation and hence the severity of the ischemic injury. Additional ECS swelling and possible membrane fragmentation may lead to further ADA reduction below normal. Therefore, a decline in ADA most probably reflects the process of cell swelling, extracellular edema, and cell lysis. The reduced ADA in phase 2 may be characterized by the development of ECS swelling and membrane degradation, while the reduced ADC can be explained by the ICS models. In contrast, persistently elevated ADA (phase 1) may suggest lack of ECS swelling and preserved membrane integrity, although vasogenic edema may be present at the early ischemic stage.⁴ The elevated ADC and reduced ADA (phase 3) most probably reflects cell lysis toward necrosis with the destruction of membrane integrity.^{4 5} Recently, reduced ADA and elevated ADC have been found in chronic white matter lesions of ischemic leukoaraiosis, which is consistent with axonal loss and gliosis.⁴¹

In patients with acute stroke observed in this study, the duration of the initial ADC reduction ranged from <48 hours to 85 days. The observed differences in ADC and ADA evolution support that ADC heterogeneity within the lesion reflects different temporal rates of stroke progression.^{5 27} The feature of a lesion with elevated ADC in the infarct core and a rim of reduced ADC at a later stage (Figure 4) has been reported in animal model studies^{3 4} with histopathological correlates of later development of infarction in the rim.⁴ In this study, the elevated ADA in the rim supports later development of ischemic infarction rather than edema. At outcome, the ADC in the rim is only slightly elevated in comparison with the infarct core, which may suggest different pathological processes. Furthermore, the outcome ADC elevation of the rim is similar to that of the small cortical GM lesions (Figure 2, bottom row). Although the measurement of ADC in small lesions or the rim area of larger lesions may be influenced by PVE, the slow recovery of the ADC and ADA in the WM lesion case (Figure 6) is less likely due to PVE. This may be a human case of reversible focal ischemic injury⁴² or ischemic-induced spreading depression found in animal models.⁴³ On the other hand, the slow evolution of ADC and ADA in these cases (in situations of better collateral flow or early reperfusion) may relate to other processes such as apoptosis, in which selective delayed cell death is likely to occur in areas of milder ischemic injury for days after the initial insult.⁴⁴

The correlation of ADA changes within 12 hours with the acute, subacute, and outcome clinical scores supports that early ADA changes reflect severity of the ischemic injury and predict stroke outcome. In contrast, the lack of correlation between clinical scores and ADA changes at later stages suggests that other processes, such as ECS swelling in vasogenic edema and cell lysis with membrane fragmentation, may have developed and contributed to a heterogeneous evolution. Furthermore, no significant correlation between ADC and clinical scores was found, suggesting that ADC alone does not have sufficient predictive power in terms of histopathological and clinical outcome in patients with acute ischemic stroke.^{5 19} This contrasts with a study in which ADC measured within 60 hours of stroke onset correlated with stroke outcome at 4 months.⁴⁵ Such disparity may relate to different sampling methods. The other study reported systematically higher ADC values by measuring the entire lesion slice, including the peripheral area.⁴⁵ This may become particularly important for small lesions, in which a measured higher ADC value may be coupled with more PVE and hence better outcome biased by the small lesion size. Second, it ignored the heterogeneity of ADC within the lesion, which may relate to different pathophysiological properties, particularly between the infarct core and periphery.^{3 4 5 27} Different processes within the lesion may already have developed within 60 hours of stroke onset. Nevertheless, whether there is an ADC threshold in predicting tissue viability and stroke outcome requires further understanding of the mechanisms underlying the ADC changes in acute ischemia. However, ADA may be the better diffusion property to track stroke progression.

Recently, elevated ADA in the acute "ischemic penumbra" delineated by DWI and CEPI has been observed and related to tissue salvage.⁴⁶ This may reflect possible membrane changes related to autoregulation of local blood flow in acute ischemia. It is possible that the ADA changes in ischemia may be influenced by early reperfusion. This is beyond the scope of the present study. Further studies to correlate the ADA changes with perfusion parameters and MRA results may help to elucidate the underlying pathophysiology in acute ischemia.

It should be noted that the empirical relationship between ADA_fr and ADA_sd was not expected theoretically. For an individual cell or fiber structure, the ADA_sd value is variable, depending on specific orientation of the cell relative to the diffusion-encoding directions. For a macroscopic ROI containing a large amount of differently orientated cells, it is possible that the average effect may reduce the orientation-dependent influence. In addition, the use of relative ratios comparing the same tissue types in the same subject may further minimize this influence and contribute to the observed correlation between ADA_fr and ADA_sd* ratios. However, the empirical relationship observed in this study may not always be valid and should not be generalized. This may be subject to further evaluation with more available DTI data.

ADA, as another diffusion property of water molecules provided by DWI, is readily available to provide additional insight into the cellular changes in acute ischemia. Combined ADC and ADA data are valuable in characterizing stroke evolution and predicting clinical outcome. These data may also help to monitor cellular changes of acute ischemia in response to putative drug treatments.

	Acknowledgments

 
This work was supported in part by the National Health and Medical Research Council, National Stroke Foundation, Lord Mayors Fund (Melbourne), Schering P/L, Neurological Foundation of New Zealand, and VJ Chapman Research Fellowship (Dr Barber). Previous funding support and continuous proprietary research support from GE Medical System are also acknowledged (Dr Yang).

Received July 1, 1999; revision received August 18, 1999; accepted August 18, 1999.

	References

Top Abstract Introduction Subjects and Methods Results Discussion References

 
1. Moseley ME, Kucharczyk J, Mintorovitch J, Cohen Y, Kurhanewicz J, Derugin N, Asgari H, Norman D. 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]

2. Knight RA, Ordidge RJ, Helpern JA, Chopp M, Rodolosi LC, Peck D. Temporal evolution of ischemic damage in rat brain measured by proton magnetic resonance imaging. Stroke. 1991;22:802–808.[Abstract/Free Full Text]

3. Reith W, Hasegawa Y, Latour LL, Dardzinski BJ, Sotak CH, Fisher M. Multislice diffusion mapping for 3-D evolution of cerebral ischemia in a rat stroke model. Neurology. 1995;45:172–177.[Abstract/Free Full Text]

4. Pierpoli C, Righini A, Linfante I, Tao-Cheng JH, Alger JR, Di Chiro G. Histopathologic correlates of abnormal water diffusion in cerebral ischemia: diffusion-weighted MR imaging and light and electron microscopic study. Radiology. 1993;189:439–448.[Abstract/Free Full Text]

5. Welch KMA, Windham J, Knight RA, Nagesh V, Hugg JW, Jacobs M, Peck D, Booker P, Dereski MO, Levine SR. A model to predict the histopathology of human stroke using diffusion and T2-weighted magnetic resonance imaging. Stroke. 1995;26:1983–1989.[Abstract/Free Full Text]

6. Warach S, Chien D, Li W, Ronthal M, Edelman RR. Fast magnetic resonance diffusion-weighted imaging of acute human stroke. Neurology. 1992;42:1717–1723.[Abstract/Free Full Text]

7. Warach S, Gaa J, Siewert B, Wielopolski P, Edelman RR. Acute human stroke studied by whole brain echo planar diffusion weighted MRI. Ann Neurol. 1995;37:231–241.[Medline] [Order article via Infotrieve]

8. Hossmann KA, Fisher M, Bockhorst K, Hoehn-Berlage M. NMR imaging of the apparent diffusion coefficient (ADC) for the evaluation of metabolic suppression and recovery after prolonged cerebral ischemia. J Cereb Blood Flow Metab. 1994;14:723–731.[Medline] [Order article via Infotrieve]

9. Sorenson AG, Buonanno FS, Gonzalez RG, Schwamm LH, Lev MH, Huang-Hellinger FR, Reese TG, Weisskoff RM, Davis TL, Suwanwela N, Can U, Moreira JA, Copen WA, Look RB, Finklestein SP, Rosen BR, Koroshetz WJ. Hyperacute stroke: evaluation with combined multisection diffusion-weighted and hemodynamically weighted echo-planar MR imaging. Radiology. 1996;199:391–401.[Abstract/Free Full Text]

10. Barber PA, Darby DG, Desmond PM, Yang Q, Gerraty RP, Jolley D, Donnan GA, Tress BM, Davis SM. Prediction of stroke outcome with echoplanar perfusion- and diffusion-weighted MRI. Neurology. 1998;51:418–426.[Abstract/Free Full Text]

11. LeBihan 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]

12. Helpern JA, Ordidge RJ, Knight RA. The effect of cell membrane water permeability on the diffusion coefficient of water. In: Proceedings of XI Annual Meeting of the Society of Magnetic Resonance in Medicine; August 8–14, 1992; Berlin, Germany. 1992;1:1201.

13. Wick M, Nagatomo Y, Prielmeier F, Frahm J. Alteration of intracellular metabolite diffusion in rat brain in vivo during ischemia and reperfusion. Stroke. 1995;26:1930–1934.[Abstract/Free Full Text]

14. van der Toorn A, Dijkhuizen RM, Tulleken CAF, Nicolay K. Diffusion of metabolites in normal and ischemic rat brain measured by localized ¹H MRS. Magn Reson Med. 1996;36:914–922.[Medline] [Order article via Infotrieve]

15. Duong TQ, Ackerman JJH, Ying HS, Neil JJ. Evaluation of extra- and intracellular apparent diffusion in normal and globally ischemic rat brain via ¹⁹F NMR. Magn Reson Med. 1998;40:1–13.[Medline] [Order article via Infotrieve]

16. van der Toorn A, Sykova E, Dijkhuizen RM, Vorisek I, Vargova L, Skobisova E, Campagne ML, Reese T, Nicolay K. Dynamic changes in water ADC, energy metabolism, extracellular space volume, and tortuosity in neonatal rat brain during global ischemia. Magn Reson Med. 1996;36:52–60.[Medline] [Order article via Infotrieve]

17. Ford JC, Hackney DB, Lavi E, Phillips M, Patel U. Dependence of apparent diffusion coefficients on axonal spacing, membrane permeability, and diffusion time in spinal cord white matter. J Magn Reson Imaging. 1998;8:775–782.[Medline] [Order article via Infotrieve]

18. Chien D, Kwong KK, Gress DR, Buonanno FS, Buxton RB, Rosen BR. MR diffusion imaging of cerebral infarction in humans. AJNR Am J Neuroradiol. 1992;13:1097–1102.[Abstract]

19. Warach S, Moseley M, Sorensen AG, Koroshetz W. Time course of diffusion imaging abnormalities in human stroke [letter]; Welch KMA, Levine SR, Chopp M, Knight RA, D'Olhaberriague L, Boska MD, Nagesh V, Windham JP, Peck D. [response]. Stroke. 1996;27:1254–1256.

20. Ulug AM, Beauchamp N Jr, Bryan RN, van Zijl PC. Absolute quantitation of diffusion constants in human stroke. Stroke. 1997;28:483–490.[Abstract/Free Full Text]

21. Warach S, Boska M, Welch KMA. Pitfalls and potential of clinical diffusion-weighted MR imaging in acute stroke. Stroke. 1997;28:481–482.

22. Pierpaoli CJ, Basser PJ. Toward a quantitative assessment of diffusion anisotropy [published correction appears in Magn Reson Med. 1997;37:972]. Magn Reson Med. 1996;36:893–906.

23. Basser PJ, Pierpaoli CJ. Microstructural and physiological features of tissues elucidated by quantitative-diffusion tensor MRI. J Magn Reson. 1996;B111:209–219.

24. van Gelderen P, de Vleeschouwer MHM, DesPres D, Pekar J, van Zijl PCM, Moonen CTW. Water diffusion and acute stroke. Magn Reson Med. 1994;31:154–163.[Medline] [Order article via Infotrieve]

25. Schlaug G, Siewert B, Benfield A, Edelman RR, Warach S. Time course of the apparent diffusion coefficient (ADC) abnormality in human stroke. Neurology. 1997;49:113–119.[Abstract/Free Full Text]

26. Schwamm LH, Koroshetz WJ, Sorensen AG, Wang B, Copen WA, Budzik R, Rordorf G, Buonanno FS, Schaefer PW, Gonzalez RG. Time course of lesion development in patients with acute stroke: serial diffusion- and hemodynamic-weighted magnetic resonance imaging. Stroke. 1998;29:2268–2276.[Abstract/Free Full Text]

27. Nagesh V, Welch KMA, Windham JP, Patel S, Levine SR, Hearshen D, Peck D, Robbins K, D'Olhaberriague L, Soltanian-Zadeh H, Boska MD. Time course of ADC_w changes in ischemic stroke: beyond the human eye! Stroke. 1998;29:1778–1782.[Abstract/Free Full Text]

28. Sykova E, Svoboda J, Polak J, Chvatal A. Extracellular volume fraction and diffusion characteristics during progressive ischemia and terminal anoxia in the spinal cord of the rat. J Cereb Blood Flow Metab. 1994;14:301–311.[Medline] [Order article via Infotrieve]

29. Thornton JS, Ordidge RJ, Penrice J, Cady EB, Amess PN, Punwani S, Clemence M, Wyatt JS. Anisotropic water diffusion in white and gray matter of the neonatal piglet brain before and after transient hypoxia-ischemia. Magn Reson Imaging. 1997;15:433–440.[Medline] [Order article via Infotrieve]

30. Carano RAD, Li F, Irie K, Fisher M, Sotak CH. The temporal evolution of diffusion anisotropy in the ischemic rat brain. In: Proceedings of the Sixth Scientific Meeting of the International Society of Magnetic Resonance in Medicine; April 18–24, 1998; Sydney, Australia; p 530.

31. Armitage PA, Bastin ME, Marshall I, Wardlaw JM, Cannon J. Diffusion anisotropy measurements in ischaemic stroke of the human brain. MAGMA. 1998;6:28–36.

32. Yang Q, Desmond PM, Tress BM, Barber PA, Darby DG, Gerraty RP, Davis SM. Apparent diffusion anisotropy in normal and ischemic human brain. In: Proceedings of the Sixth Scientific Meeting of the International Society of Magnetic Resonance in Medicine; April 18–24, 1998; Sydney, Australia; p 1250.

33. Zelaya F, Flood N, Chalk JB, Wang D, Strugnell W, Doddrell DM, Strugnell W, Benson M, Ostergaard L, Semple J, Eagle S. 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.[Medline] [Order article via Infotrieve]

34. Cote R, Battista RN, Wolfson C, Boucher J, Adam J, Hachinski V. The Canadian Neurological Scale: validation and reliability assessment. Neurology. 1989;39:638–643.[Abstract/Free Full Text]

35. Wade DT. Measurement in Neurological Rehabilitation. Oxford, England: Oxford Medical Publications; 1994.

36. Stejskal EO, Tanner JE. Spin diffusion measurements: spin-echoes in the presence of a time-dependent field gradient. J Chem Physiol. 1965;42:288–292.

37. Haselgrove JC, Moore JR. Correction for distortion of echo-planar images used to calculate the apparent diffusion coefficient. Magn Reson Med. 1996;36:960–964.[Medline] [Order article via Infotrieve]

38. Park S, Schowengerdt R. Image reconstruction by parametric cubic convolution. Comput Vis Graph Image Proc. 1983;23:258–272.

39. Siesjo BK, Bengtsson F. Calcium fluxes, calcium antagonists, and calcium-related pathology in brain ischemia, hypoglycemia, and spreading depression: a unifying hypothesis. J Cereb Blood Flow Metab. 1989;9:127–140.[Medline] [Order article via Infotrieve]

40. Loubinoux I, Volk A, Borredon J, Guirimand S, Tiffon B, Seylaz J. Spreading of vasogenic edema and cytotoxic edema assessed by quantitative diffusion and T2 magnetic resonance imaging. Stroke. 1997;28:419–427.[Abstract/Free Full Text]

41. Jones DK, Lythgoe D, Mark A, Horsfield MA, Simons A, Williams SCR, Markus HS. Characterization of white matter damage in ischemic leukoaraiosis with diffusion tensor MRI. Stroke. 1999;30:393–397.[Abstract/Free Full Text]

42. Minematsu K, Li L, Sotak CH, Davis MA, Fisher M. Reversible focal ischemic injury demonstrated by diffusion-weighted magnetic resonance imaging in rats. Stroke. 1992;23:1304–1311.[Abstract/Free Full Text]

43. Rother J, de Crespigny AJ, D'Arceuil H, Iwai K, Moseley ME. Recovery of apparent diffusion coefficient after ischemia-induced spreading depression relates to cerebral perfusion gradient. Stroke. 1996;27:980–987.[Abstract/Free Full Text]

44. Pulera MR, Adams LM, Liu H, Santos DG, Nishimura RN, Yang F, Cole GM, Wasterlain CG. Apoptosis in a neonatal rat model of cerebral hypoxia-ischemia. Stroke. 1998;29:2622–2630.[Abstract/Free Full Text]

45. van Everdingen KJ, van der Grond J, Kappelle LJ, Ramos LMP, Mali WPTM. Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke. 1998;29:1783–1790.[Abstract/Free Full Text]

46. Yang Q, Tress BM, Barber PA, Li T, Darby DG, Gerraty RP, Desmond PM, Davis SM. Prognosis of tissue viability in human stroke using diffusion and perfusion imaging. In: Proceedings of the Annual Meeting of the American Society of Neuroradiology; May 23–28, 1999; San Diego, Calif; p 119.

This article has been cited by other articles:

				K. Butcher and C. Beaulieu Physiological clocks in acute ischaemic stroke: the search continues J. Neurol. Neurosurg. Psychiatry, September 1, 2009; 80(9): 941 - 941. [Full Text] [PDF]

				K Sakai, K Yamada, Y Nagakane, S Mori, M Nakagawa, and T Nishimura Diffusion tensor imaging may help the determination of time at onset in cerebral ischaemia J. Neurol. Neurosurg. Psychiatry, September 1, 2009; 80(9): 986 - 990. [Abstract] [Full Text] [PDF]

				Y. Kusano, T. Seguchi, T. Horiuchi, Y. Kakizawa, T. Kobayashi, Y. Tanaka, K. Seguchi, and K. Hongo Prediction of Functional Outcome in Acute Cerebral Hemorrhage Using Diffusion Tensor Imaging at 3T: A Prospective Study AJNR Am. J. Neuroradiol., September 1, 2009; 30(8): 1561 - 1565. [Abstract] [Full Text] [PDF]

				M. M. L. de Win, G. Jager, J. Booij, L. Reneman, T. Schilt, C. Lavini, S. D. Olabarriaga, G. J. den Heeten, and W. van den Brink Sustained effects of ecstasy on the human brain: a prospective neuroimaging study in novel users Brain, November 1, 2008; 131(11): 2936 - 2945. [Abstract] [Full Text] [PDF]

				M.L. Ishimori, B.D. Pressman, D.J. Wallace, and M.H. Weisman Posterior reversible encephalopathy syndrome: another manifestation of CNS SLE? Lupus, June 1, 2007; 16(6): 436 - 443. [Abstract] [PDF]

				Y. Liu, H. E. D'Arceuil, S. Westmoreland, J. He, M. Duggan, R. G. Gonzalez, J. Pryor, and A. J. d. Crespigny Serial Diffusion Tensor MRI After Transient and Permanent Cerebral Ischemia in Nonhuman Primates Stroke, January 1, 2007; 38(1): 138 - 145. [Abstract] [Full Text] [PDF]

				C. van Pul, J. Buijs, M. J. A. Janssen, G. F. Roos, M. T. Vlaardingerbroek, and P. F. F. Wijn Selecting the Best Index for Following the Temporal Evolution of Apparent Diffusion Coefficient and Diffusion Anisotropy After Hypoxic-Ischemic White Matter Injury in Neonates AJNR Am. J. Neuroradiol., March 1, 2005; 26(3): 469 - 481. [Abstract] [Full Text] [PDF]

				J. Konishi, K. Yamada, O. Kizu, H. Ito, K. Sugimura, K. Yoshikawa, M. Nakagawa, and T. Nishimura MR tractography for the evaluation of functional recovery from lenticulostriate infarcts Neurology, January 11, 2005; 64(1): 108 - 113. [Abstract] [Full Text] [PDF]

				S M. Maniega, M E Bastin, P A Armitage, A J Farrall, T K Carpenter, P J Hand, V Cvoro, C S Rivers, and J M Wardlaw Temporal evolution of water diffusion parameters is different in grey and white matter in human ischaemic stroke J. Neurol. Neurosurg. Psychiatry, December 1, 2004; 75(12): 1714 - 1718. [Abstract] [Full Text] [PDF]

				Y. Ozsunar, P. E. Grant, T. A. G. M. Huisman, P. W. Schaefer, O. Wu, A. G. Sorensen, W. J. Koroshetz, and R. G. Gonzalez Evolution of Water Diffusion and Anisotropy in Hyperacute Stroke: Significant Correlation between Fractional Anisotropy and T2 AJNR Am. J. Neuroradiol., May 1, 2004; 25(5): 699 - 705. [Abstract] [Full Text] [PDF]

				J. L. Bykowski, L. L. Latour, and S. Warach More Accurate Identification of Reversible Ischemic Injury in Human Stroke by Cerebrospinal Fluid Suppressed Diffusion-Weighted Imaging Stroke, May 1, 2004; 35(5): 1100 - 1106. [Abstract] [Full Text] [PDF]

				T. A.G.M. Huisman, L. H. Schwamm, P. W. Schaefer, W. J. Koroshetz, N. Shetty-Alva, Y. Ozsunar, O. Wu, and A. G. Sorensen Diffusion Tensor Imaging as Potential Biomarker of White Matter Injury in Diffuse Axonal Injury AJNR Am. J. Neuroradiol., March 1, 2004; 25(3): 370 - 376. [Abstract] [Full Text] [PDF]

				J. R. Lynch, R. Blessing, W. D. White, H. P. Grocott, M. F. Newman, and D. T. Laskowitz Novel Diagnostic Test for Acute Stroke Stroke, January 1, 2004; 35(1): 57 - 63. [Abstract] [Full Text] [PDF]

				K. Yamada, S. Mori, H. Nakamura, H. Ito, O. Kizu, K. Shiga, K. Yoshikawa, M. Makino, S. Yuen, T. Kubota, et al. Fiber-Tracking Method Reveals Sensorimotor Pathway Involvement in Stroke Patients Stroke, September 1, 2003; 34 (9): e159 - e162. [Abstract] [Full Text] [PDF]

				A. S. Field, K. Hasan, B. J. Jellison, K. Arfanakis, and A. L. Alexander Diffusion Tensor Imaging in an Infant with Traumatic Brain Swelling AJNR Am. J. Neuroradiol., August 1, 2003; 24(7): 1461 - 1464. [Abstract] [Full Text] [PDF]

				K Yamada, O Kizu, H Ito, H Nakamura, S Yuen, K Yoshikawa, K Shiga, and T Nishimura Wallerian degeneration of the inferior cerebellar peduncle depicted by diffusion weighted imaging J. Neurol. Neurosurg. Psychiatry, July 1, 2003; 74(7): 977 - 978. [Abstract] [Full Text] [PDF]

				J. D. Eastwood, S. T. Engelter, J. F. MacFall, D. M. Delong, and J. M. Provenzale Quantitative Assessment of the Time Course of Infarct Signal Intensity on Diffusion-Weighted Images AJNR Am. J. Neuroradiol., April 1, 2003; 24(4): 680 - 687. [Abstract] [Full Text] [PDF]

				P. W. Schaefer, Y. Ozsunar, J. He, L. M. Hamberg, G. J. Hunter, A. G. Sorensen, W. J. Koroshetz, and R. G. Gonzalez Assessing Tissue Viability with MR Diffusion and Perfusion Imaging AJNR Am. J. Neuroradiol., March 1, 2003; 24(3): 436 - 443. [Abstract] [Full Text] [PDF]

				J.M. Wardlaw, S.L. Keir, M.E. Bastin, P.A. Armitage, and A.K. Rana Is diffusion imaging appearance an independent predictor of outcome after ischemic stroke? Neurology, November 12, 2002; 59(9): 1381 - 1387. [Abstract] [Full Text] [PDF]

				J. P. Broderick and W. Hacke Treatment of Acute Ischemic Stroke: Part I: Recanalization Strategies Circulation, September 17, 2002; 106(12): 1563 - 1569. [Full Text] [PDF]

				H. A.L. Green, A. Pena, C. J. Price, E. A. Warburton, J. D. Pickard, T. A. Carpenter, and J. H. Gillard Increased Anisotropy in Acute Stroke: A Possible Explanation Stroke, June 1, 2002; 33(6): 1517 - 1521. [Abstract] [Full Text] [PDF]

				C. Oppenheim, C. Grandin, Y. Samson, A. Smith, T. Duprez, C. Marsault, and G. Cosnard Is There an Apparent Diffusion Coefficient Threshold in Predicting Tissue Viability in Hyperacute Stroke? Stroke, November 1, 2001; 32(11): 2486 - 2491. [Abstract] [Full Text] [PDF]

				W. A. Copen, L. H. Schwamm, R. G. Gonzalez, O. Wu, C. B. Harmath, P. W. Schaefer, W. J. Koroshetz, and A. G. Sorensen Ischemic Stroke: Effects of Etiology and Patient Age on the Time Course of the Core Apparent Diffusion Coefficient Radiology, October 1, 2001; 221(1): 27 - 34. [Abstract] [Full Text] [PDF]

				P. M. Desmond, A. C. Lovell, A. A. Rawlinson, M. W. Parsons, P. A. Barber, Q. Yang, T. Li, D. G. Darby, R. P. Gerraty, S. M. Davis, et al. The Value of Apparent Diffusion Coefficient Maps in Early Cerebral Ischemia AJNR Am. J. Neuroradiol., August 1, 2001; 22(7): 1260 - 1267. [Abstract] [Full Text] [PDF]

				M. W. Parsons, Q. Yang, P. A. Barber, D. G. Darby, P. M. Desmond, R. P. Gerraty, B. M. Tress, and S. M. Davis Perfusion Magnetic Resonance Imaging Maps in Hyperacute Stroke : Relative Cerebral Blood Flow Most Accurately Identifies Tissue Destined to Infarct Stroke, July 1, 2001; 32(7): 1581 - 1587. [Abstract] [Full Text] [PDF]

				S. Higano, J. Zhong, D. A. Shrier, D. K. Shibata, Y. Takase, H. Wang, and Y. Numaguchi 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., March 1, 2001; 22(3): 456 - 463. [Abstract] [Full Text]

				S. L. Keir and J. M. Wardlaw Systematic Review of Diffusion and Perfusion Imaging in Acute Ischemic Stroke Stroke, November 1, 2000; 31(11): 2723 - 2731. [Abstract] [Full Text] [PDF]

				D. C. Tong, A. Adami, M. E. Moseley, and M. P. Marks Relationship Between Apparent Diffusion Coefficient and Subsequent Hemorrhagic Transformation Following Acute Ischemic Stroke Stroke, October 1, 2000; 31(10): 2378 - 2384. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yang, Q. Articles by Davis, S. M. Search for Related Content PubMed PubMed Citation Articles by Yang, Q. Articles by Davis, S. M.