ORIGINAL ARTICLE

False-Negative Diffusion-Weighted Imaging in Acute Stroke and its Frequency in Anterior and Posterior Circulation Ischemia Haci Taner Bulut, MD,* Adem Yildirim, MD,† Burcu Ekmekci, MD,‡ Neslihan Eskut, MD,§ and Hediye Pinar Gunbey, MD∥ Objective: We aimed to investigate the location and size of ischemic stroke lesions that were frequently overlooked by diffusion-weighted imaging (DWI). Materials and Methods: We retrospectively reviewed the medical records of 162 patients who had symptoms suggesting ischemic stroke. National Institutes of Health Stroke Scale and Modified Rankin Scale scores, lesion size, magnetic resonance imaging (MRI) findings, delay between onset of symptoms and initial MRI (MRI latency), and vascular distribution of the stroke lesions were analyzed in patients with falsenegative DWI findings. Results: Of the 116 patients with a final diagnosis of acute ischemic stroke, 11 patients (9.48%) had false-negative DWI findings in the initial period. The mean (SD) MRI latency was 4.3 (1.2) hours. There was no statistically significant difference in point of lesion size, the National Institutes of Health Stroke Scale, and the Modified Rankin Scales scores. Conclusions: False-negative DWI findings in acute stroke can be observed both in association with the posterior circulation/small lesions and the anterior circulation/large lesions. Key Words: stroke, diffusion-weighted imaging, false-negative, magnetic resonance imaging (J Comput Assist Tomogr 2014;38: 627–633)

D

iffusion-weighted imaging (DWI) is the most frequently used technique for the diagnosis of acute ischemic stroke. It has been proposed to be superior to other methods such as computed tomography (CT) or conventional magnetic resonance imaging (MRI) for examining patients presenting with acute ischemic stroke.1–7 Animal studies have shown that changes on DWI can be determined within 1 hour after middle cerebral artery occlusion and on apparent diffusion coefficient (ADC) maps within 3 minutes after global occlusion.8,9 However, numerous reports on diffusion-negative stroke illustrate that the diagnosis of acute ischemic stroke cannot be excluded solely based on the absence of lesions on DWI.10–15 In these studies, the rate of false-negative DWI results ranged from 5.8% to 19%.1,11 Hence, the reliability of initial negative DWI studies for excluding ischemic stroke seems doubtful. In addition, most previous studies have found that posterior circulation and small lacunar or subcortical lesions are associated with false-negative DWI results.1,3,4,10,11,16 Few studies have reported that nonlacunar

From the Departments of *Radiology, †Physical Medicine and Rehabilitation, and ‡Neurology, Medical Faculty of Adiyaman University, Adiyaman, Turkey; §Department of Neurology, Adiyaman Education and Research Hospital, Adiyaman, Turkey; and ∥Department of Radiology, Samsun Education and Research Hospital, Samsun, Turkey. Received for publication December 22, 2013; accepted March 18, 2014. Reprints: Haci Taner Bulut, MD, Department of Radiology, Medical Faculty of Adiyaman University, Turkey (e‐mail: [email protected]). The authors declare no conflicts of interest. Copyright © 2014 by Lippincott Williams & Wilkins

anterior circulatory lesions may also be associated with falsenegative DWI results.12–14 New initial false-negative DWI studies that depict large lesions, which are located in both anterior and posterior circulatory due to nonsmall infarcts, are needed; and this study is focused on this issue. At our hospital, DWI is used as a first-line diagnostic tool for patients admitted for a suspected ischemic event based on the sudden onset of neurological symptoms. In this study, we investigated the location and size of ischemic stroke lesions that were frequently overlooked by DWI as well as the rate of falsenegative DWI results in patients with permanent neurological deficits due to acute ischemic stroke.

MATERIALS AND METHODS Patients This study was approved by the ethics committee of our university. We retrospectively evaluated 162 patients between November 2011 and July 2013 who were admitted to the hospital with signs and symptoms highly suggestive of arterial ischemic stroke and who eventually received a diagnosis of acute ischemic stroke for which the time of onset could be determined. The exclusion criteria were a baseline CT scan with evidence of hemorrhage and stroke mimics as a cause of the initial negative DWI with an undetermined time of onset and artifacts present on MRI. Among the 162 initially evaluated patients, 116 patients were included in the study. The remaining 46 patients had a final diagnosis other than arterial stroke: 22 patients had evidence of hemorrhage on their CT scans, 2 patients had venous infarcts, 3 patients had stroke mimics as a cause of the initial negative DWI, 14 patients had an undetermined time of onset of disease, and 5 patients had artifacts present on MRI. All available clinical data, paraclinical studies, including etiological cardiac and arterial workup, and radiological follow-up, were reviewed by a stroke neurologist. The National Institute of Health Stroke Scale (NIHSS) score on admission and the modified Rankin Scale (mRS) score upon discharge were recorded for all patients.

Magnetic Resonance Imaging Magnetic resonance imaging was performed with a 1.5-T MR unit equipped with echo-planar capability (Magnetom Syngo Symphony; Siemens Medical Systems, Erlangen, Germany). Before MR imaging, all patients underwent an unenhanced CT scan to exclude intracranial hemorrhage. All MRI studies were performed within 30 to 45 minutes after admission to the hospital. Axial DWI images were recorded using a single-shot spin-echo echoplanar pulse sequence (field of view, 24 cm; thickness of sections, 5 mm; matrix, 128  100; repetition time (TR)/echo time (TE), 4821/106 milliseconds (ms) ; b values, 0-1000 seconds/mm2). Apparent diffusion coefficient maps were automatically calculated by manufacturer’s software program. Coronal fluid-attenuated

J Comput Assist Tomogr • Volume 38, Number 5, September/October 2014

www.jcat.org

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

627

628

www.jcat.org

No RF

DM HT, DM

HT, DM

No RF

CAD

HT

CAD

CAD

No RF

41/M

63/M 50/M

75/M

68/F

57/F

78/M

81/F

82/M

82/F

70/F

Case No

1

2 3

4

5

6

7

8

9

10

11

Symptoms

Dysarthria, right extremity ataxia Left HP Right ptosis, dysphagia, left HP Left CFP and left HP Right HP

4

6

2

11

4

13

2

5

2 2

4

NIHSS Score at Admission

Left cortical-subcortical parietal lobe Brain stem

Right cortical paracentral lobule Left basal ganglia and corona radiata Left frontoparietal lobe and basal ganglia Left cortical paracentral lobule Left cortical-subcortical parietal lobe Right Occipital lobe

Right cerebellum and brain stem Brain stem Brain stem

Abnormal Anatomic Region on MRI

2

1

1

3

2

6

1

2

1 2

2

5

4

3

4

7

4

3

5

5 3

4

34

37

29

19

30

17

34

18

10 24

25

Time Time From Ranking From Symptom Score at Symptom Onset to Hospital Onset to Follow-Up Discharge MRI (h) MRI (h) Right VA occlusion

Color Duplex Examination

NS

NS

NS

NS

NS

NS

NS

0.068 Right VA occlusion

1.4

14.9

0.715

0.840

10.2

1.2

3.6

0.084 Right VA occlusion 0.25 Right VA occlusion

5.55

Size of Lesion (cm3)

MRA

Right VA occlusion

NP

Right PCA OB occlusion

NP

NS

NP

NS

NP

Right VA occlusion Right VA occlusion

Right VA occlusion

CAD, coronary artery disease; CFP, central facial paresis; DM, diabetes mellitus; HP, hemiparesis; HT, hypertension; MA, motor aphasia; NP, not performed; NS, nonsignificant stenosis; OB, occipital branch; PCA, posterior cerebral artery; RF, risk factor; SA, sensorial aphasia; SMA, sensory motor aphasia; VA, vertebral artery.

Right HP

Homonymous hemianopsia SA

SMA and right HP

SMA and right HP

HT, DM, CAD Lethargia and right HP

Stroke Risk Factor

Age (y)/Sex

TABLE 1. Findings in 11 Patients With Acute Ischemic Stroke and False-Negative DWI

Bulut et al J Comput Assist Tomogr • Volume 38, Number 5, September/October 2014

© 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

J Comput Assist Tomogr • Volume 38, Number 5, September/October 2014

inversion recovery (FLAIR) images were obtained with a 24-cm viewing field and no-gap sections (5 mm thick) with a 256  256 matrix, 9002/116 ms TR/TE, and 2500-ms inversion time. Conventional spin-echo axial T1-weighted MR images were obtained with a 24-cm viewing field and sections (5 mm thick) with a 1-mm gap 501/8.2-ms TR/TE. Axial turbo spin echo T2weighted images were obtained with a 24-cm viewing field, sections (5 mm thick) with a 1-mm gap, 320  162 matrix, and 5382/125-ms TR/TE. Initial false-negative DWI studies corresponded to normal initial DWI findings, a stroke lesion clearly visible on followup MR studies, and a final diagnosis of arterial stroke. All initial MR images were retrospectively reviewed by a neuroradiologist who was informed of the acute neurological symptoms, but not of the clinical and imaging follow-ups. The neuroradiologist accepted the criteria that hyperintensities on DWI with reduced (hypointense) ADC were seen in clinically associated brain regions for the diagnosis of arterial stroke. In light of this, the review was compared to the report made at the time of initial MRI. A follow-up MR study, including T1-weighted, T2-weighted, FLAIR, and DWI, was systematically performed on patients with negative initial DWI findings before hospital discharge. Six months after the initial scan, a follow-up MRI was performed on the 3 patients with negative initial DWI findings.

Statistical Method Statistical analyses were performed using SPSS software (Statistical Package for the Social Sciences, version 18; SPSS, Inc, Chicago, Ill). The clinical data, lesion size, MRI findings, delay between onset of stroke symptoms to initial MRI (MRI latency), and vascular distribution of the stroke lesions in patients with false-negative DWI findings were analyzed. The falsenegative and positive initial DWI results with regard to vascular distribution, NIHSS, and mRS scores of patients were compared and statistically analyzed. The independent-samples t test was performed to compare NIHSS and mRS scores, and the χ2 test was used to study the vascular distribution. The influence of MRI latency on the probability of initial positive DWI findings was analyzed using logistic regression: the probability of initial positive DWI was graphically plotted against time (MRI latency). Lesion size was estimated from follow-up coronal FLAIR images. The brain lesion area consistent with clinical ischemia was manually plotted on each image using an image analysis program tool powered by Siemens. The area of abnormality on each section was multiplied by the section thickness to obtain the volume. For all statistical analyses, a difference was considered significant at P < 0.05.

RESULTS Of the 116 patients with a final diagnosis of acute ischemic stroke, 11 patients (9.48%) had false-negative DWI findings in the initial period. The mean (SD) age of all patients was 71.5 (11.3) (range, 22.0–113.0) years; 60 patients were men and 56 patients were women. Table 1 summarizes the demographic characteristics, stroke risk factors, symptoms, NIHSS scores on admission, locations of lesions, mRS scores upon hospital discharge, delay between onset of stroke symptoms to initial MRI and follow-up MRI, size of the lesions, and colored Doppler ultrasonography and MR angiography findings of the 11 patients with initial false-negative DWI findings. Acute ischemic stroke areas with clinically relevant brain regions were seen on control DWI, T2-weighted, and FLAIR images in patients with initial negative DWI (Figs. 1–3). Initial FLAIR imaging showed positive findings in 3 patients with © 2014 Lippincott Williams & Wilkins

False-Negative DW Imaging in Acute Stroke

initial negative DWI (Fig. 4). In 94 patients (94/116 [81.0%]), the stroke lesions were located in the anterior circulation; and in 22 patients (22/116 [19.0%]), they were located in the posterior circulation. The lesions in 6 of 11 patients with falsenegative DWI were located in the anterior circulation (6/94 [6.38%]), and the remaining 5 patients’ lesions were in the posterior circulation (5/22 [22%]). Among all ischemic strokes, the false-negative DWI rate of lesions was higher in the posterior circulation (22.00%) than in the anterior circulation (6.38%), and this difference was statistically significant (χ2, 5.54; P = 0.018; 1 df ). A total of 3 of the 11 patients with initial negative DWI findings underwent 6-month control MR examinations, and encephalomalacic changes were observed in the initial ischemic areas (Figs. 1, 3). The MRI latency of the 11 patients with initial false-negative DWI findings varied between 3 and 7 hours (mean [SD], 4.3 [1.2] hours), and the time between the onset of the stroke symptoms to follow-up MRI examination of these patients varied between 10 and 37 hours (mean [SD], 25.18 [8.49] hours). The MRI latency of the 105 patients with initial positive DWI findings varied between 2 and 48 hours (mean [SD], 10.79 [10.26] hours) as follows: 0 to 4.5 hours, 25 patients; 4.5 to 6 hours, 24 patients; 6 to 12 hours, 37 patients; 12 to 24 hours, 10 patients; and 24 to 48 hours, 9 patients. The results of the logistic regression analysis for MRI latency on the probability of initial positive DWI findings are presented in Figure 5. The occurrence of initial positive DWI findings increased significantly as MR latency increased (χ2 = 13.11; P = 0.0003; 1 df ). The size of the lesions in the patients with false-negative DWI findings varied between 0.068 and 14.9 cm3 (mean [SD], 3.53 [4.87]). The size of the lesions in patients with positive initial DWI findings varied between 0.076 and 21.236 cm3 (mean [SD], 4.08 [3.91]). No statistically significant difference was found between the 2 groups with regard to lesion size (P = 0.258). In the patients with false-negative DWI findings, the mean (SD) NIHSS was 5.00 (3.74); in the patients with initial positive DWI findings, the mean (SD) NIHSS was 6.17 (4.14). There was no statistically significant difference between the NIHSS scores (P = 0.370). The mean (SD) mRS score was 2.01 (1.45) in the patients with false-negative DWI findings and 2.23 (1.28) in the patients with initial positive DWI findings. There was no statistically significant difference between the mRS scores (P = 0.739).

DISCUSSION Although a high success rate has been reported for DWI in patients with ischemic stroke, a non-negligible number of articles have demonstrated that many patients with ischemic stroke symptoms can be overlooked by DWI.1–7 Similarly, our findings have demonstrated that the diagnosis of ischemic stroke cannot be ruled out in the absence of positive DWI findings, and false negativity of initial DWI in patients with acute stroke is not rare. Of the studies that mention false-negative DWI results, Oppenheim et al11 reported that DWI may lack the resolution required to show small lesions. In addition, because of the short period between the onset of stroke and imaging, DWI may not obtain sufficient signals for lesions located in the vertebral basilar system, and lesions in the brain stem might be missed because of magnetic susceptibility artifacts as possible reasons for false negativity. In another study by Chalela et al,1 3 reasons were proposed for false negativity. The first was lesions located in the brain stem, the second was a less than 3-h period between the onset of stroke symptoms and imaging, and the third was an NIHSS score of less than 4. However, they also reported that almost half of their false-negative DWI cases did not have any of the aforementioned 3 features. Wang et al12 and Lefkowitz et al13 www.jcat.org

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

629

J Comput Assist Tomogr • Volume 38, Number 5, September/October 2014

Bulut et al

FIGURE 1. Sudden-onset dysarthria and right extremity ataxia in a 41-year-old man. Axial DWI MR images (A), axial T2-weighted MR images (B), and coronal FLAIR MR images (C) show the initial (considered normal) on the left side and the follow-up MRI hyperintensities (stroke lesion) on the right side. D, Axial T2-weighted MR images show encephalomalacic stroke area (follow-up imaging 6 months after initial scan).

FIGURE 2. Sudden-onset lethargia and right hemiparesis in a 57-year-old woman. Axial DWI MR images (A), axial T2-weighted MR images (B), axial ADC map images (C), and coronal FLAIR MR images (D) show the initial (considered normal) on the left side and the follow-up MRI stroke lesion on the right side.

630

www.jcat.org

© 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

J Comput Assist Tomogr • Volume 38, Number 5, September/October 2014

False-Negative DW Imaging in Acute Stroke

FIGURE 3. Sudden-onset homonymous hemianopsia in an 82-year-old man. Axial DWI MR images (A), axial T2-weighted MR images (B), and coronal FLAIR MR images (C) show the initial (considered normal) on the left side and the follow-up MRI hyperintensities (stroke lesion) on the right side. D, Axial T2-weighted and FLAIR MR images show encephalomalacic stroke area (follow-up imaging 6 months after initial scan).

FIGURE 4. Sudden-onset lethargia and right hemiparesis in a 57-year-old woman. Axial DWI MR images (A), axial T2-weighted MR images (B), axial ADC map images (C), and coronal FLAIR and ADC map images (D) show initial positive FLAIR findings with initial negative ADC (T2 shine through) on the left side and follow-up MRI stroke lesion on the right side. © 2014 Lippincott Williams & Wilkins

www.jcat.org

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

631

J Comput Assist Tomogr • Volume 38, Number 5, September/October 2014

Bulut et al

FIGURE 5. Estimation of the probability of initial positive DWI findings by means of a logistic regression. The probability of initial positive DWI findings is plotted against time for stroke lesions. Initial positive DWI findings increases with time.

reported 2 cases of false negativity involving lesions located in the anterior circulation. They proposed that the degree of hypoperfusion in the early phase of ischemia remained below the threshold required to form an image with DWI as the reason for the false negativity. Our study suggests that perfusion defects are not severe enough to create a finding on DWI, thus resulting in false negativity. The higher rate of false-negative DWI results in the posterior circulation seems to be related to lesions located in the brain stem. For this reason, we suggest that false-negative DWI findings are basically due to lesions located in the brain stem and the lack of a sufficient perfusion defect. In ischemic stroke, the relationships among stroke, cerebral blood flow (CBF), clinical symptoms, and time were first explained in animal model studies, then in human studies. These studies demonstrated that paralysis after vascular occlusion occurs when CBF decreases below a certain threshold value. However, paralysis is not equivalent to an infarct, and if the CBF remains above threshold values, brain tissue can remain alive and can completely recover after reperfusion. It has been shown that tissues develop infarction with low blood flow, but irreversible damage occurs over a period of time, and brain tissue infarction can occur within minutes. It has also been demonstrated that with high CBF levels such as 15 to 20 mL/100 g per minute, infarction may not develop for more than 2 hours.17 Again, it has been reported that the development process of brain tissue infarction below the threshold value differs in each individual.12,15,17 When the aforementioned findings are evaluated along with the false-negative DWI results reported by some studies,12–15 it can be assumed that although there may be clinical symptoms, false-negative DWI can be seen with infarcts if a sufficient perfusion defect is not present. In our study, the initial MRI screening of patients was performed within the first few hours after the onset of symptoms (mean, 4.3 hours). We suggest that the hypoperfusion level in the first few hours after symptom onset may not be enough to cause infarction and can thus lead to false-negative findings on the initial DWI. With regard to other possible reasons for false negativity associated with lesions located in the brain stem, we believe that DWI lacks the sufficient resolution to show small lesions; they do not obtain sufficient signals on DWI because of the short period between the onset of symptoms and MRI. Artifact lesions may go undetected owing to magnetic susceptibility.

632

www.jcat.org

In our study, the patients with false-negative DWI findings had large lesions located in both the anterior and posterior circulation. There was no statistically significant difference in lesion size between the patients with initial positive and initial negative DWI results. In most previous studies, false-negative DWI findings have been associated with the posterior circulation and small lesions.1,3,4,10,11,16 Only a few studies have reported false-negative DWI findings involving the anterior circulation and large lesions.12–15 Contrary to these studies, our findings suggest that there is no relationship between false-negative DWI findings and lesion size. It is clear that the relationship found between false-negative DWI results and small lesion size in these previous studies were due to lesions located in the brain stem. Some studies,12–14 have reported large lesions with false DWI findings. Our results support the conclusion that lesion size is not related to false-negative DWI findings. To the best of our knowledge, no previous study has reported a large series of false-negative DWI findings in large lesions located in both the anterior and posterior circulation. Our results emphasize the fact that false-negative DWI findings in acute stroke not only involve the posterior circulation and small lacunar or subcortical lesions but that they can also be observed in the anterior circulation and with large lesions. Of the studies published to date, Tong et al14 reported NIHSS scores of 17 and 24 for 2 false-negative DWI cases, whereas Oppenheim et al11 reported a score of 12 in one case. In approximately half of the false-negative DWI cases in Chalela et al,1 the NIHSS score was less than 4; however, no relationship was found between false negativity and NIHSS scores in the other half of the cases. In the present study, the mean NIHSS score of the false-negative DWI patients was 5; in only 2 cases, the NIHSS scores were greater than 10 (11 and 13, respectively). We did not find any relationships between NIHSS scores and false-negative DWI results. This study has a few limitations. First, we did not perform perfusion-weighted imaging. Second, our study was retrospective, and a relatively small number of subjects were investigated. A further prospective larger study including both DWI and perfusion-weighted imaging is needed.

CONCLUSIONS False-negative DWI findings in acute stroke can be observed both in association with the posterior circulation/small lacunar or subcortical lesions and the anterior circulation/large lesions. Moreover, initial negative DWI findings are not infrequent during the first few hours of a stroke. Therefore, the diagnosis of stroke should not be excluded based on early negative DWI findings. ACKNOWLEDGMENTS The authors thank all the individuals who have participated in the study. REFERENCES 1. Chalela JA, Kidwell CS, Nentwich LM, et al. Magnetic resonance imaging and computed tomography in emergency assessment of patients with suspected acute stroke: a prospective comparison. Lancet. 2007;369:293–298. 2. Warach S, Gaa J, Siewert B, et al. Acute human stroke studied by whole brain echo planar diffusion-weighted magnetic resonance imaging. Ann Neurol. 1995;37:231–241.

© 2014 Lippincott Williams & Wilkins

Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

J Comput Assist Tomogr • Volume 38, Number 5, September/October 2014

False-Negative DW Imaging in Acute Stroke

3. Lovblad KO, Laubach HJ, Baird AE, et al. Clinical experience with diffusion-weighted MR in patients with acute stroke. Am J Neuroradiol. 1998;19:1061–1066.

10. Morita S, Suzuki M, Iizuka K. False-negative diffusion-weighted MRI in acute cerebellar stroke. Auris Nasus Larynx. 2011; 38:577–582.

4. 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.

11. Oppenheim C, Stanescu R, Dormont D, et al. False-negative diffusion-weighted MR findings in acute ischemic stroke. AJNR Am J Neuroradiol. 2000;21:1434–1440.

5. van Everdingen KJ, van der Grond J, Kappelle LJ, et al. Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke. 1998;29:1783–1790.

12. Wang PY, Barker PB, Wityk RJ, et al. Diffusion-negative stroke: a report of two cases. AJNR Am J Neuroradiol. 1999;20:1876–1880.

6. Schellinger PD, Jansen O, Fiebach JB, et al. Feasibility and practicality of MR imaging of stroke in the management of hyperacute cerebral ischemia. AJNR Am J Neuroradiol. 2000;21:1184–1189. 7. Singer MB, Chong J, Lu D, et al. Diffusion-weighted MRI in acute subcortical infarction. Stroke. 1998;29:133–136. 8. Moseley ME, Kucharczyk J, Mintorovitch J, et al. Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility-enhanced MR imaging in cats. AJNR Am J Neuroradiol. 1990;11:423–429. 9. Pierpaoli C, Alger JR, Righini A, et al. High temporal resolution diffusion MRI of global cerebral ischemia and reperfusion. J Cereb Blood Flow Metab. 1996;16:892–905.

© 2014 Lippincott Williams & Wilkins

13. Lefkowitz D, LaBenz M, Nudo SR, et al. Hyperacute ischemic stroke missed by diffusion-weighted imaging. AJNR Am J Neuroradiol. 1999;20:1871–1875. 14. Tong DC, Yenari MA, Albers GW, et al. Correlation of perfusion- and diffusion-weighted MRI with NIHSS score in acute (

False-negative diffusion-weighted imaging in acute stroke and its frequency in anterior and posterior circulation ischemia.

We aimed to investigate the location and size of ischemic stroke lesions that were frequently overlooked by diffusion-weighted imaging (DWI)...
6MB Sizes 0 Downloads 3 Views