Int J Clin Exp Med 2018;11(1):109-117www.ijcem.com /ISSN:1940-5901/IJCEM0060811

Original ArticleApparent diffusion coefficient value for prediction of hemorrhagic transformation in acute ischemic infarction

Li-Peng Chen1,2, Xin-Qing Jiang1,2, Zhao Zhao1, Xiu-He Wang1

1The First Affiliated Hospital of Jinan University, Guangzhou 510630, China; 2Guangzhou First People’s Hospital, Guangzhou Medical University, Guangzhou 510180, China

Received February 6, 2017; Accepted November 27, 2017; Epub January 15, 2018; Published January 30, 2018

Abstract: Objective: We investigated the role of apparent diffusion coefficient (ADC) values in assessing the risk of hemorrhagic transformation in acute cerebral infarcts. Materials and methods: Fifty-eight patients with acute isch-emic infarction undergoing magnetic resonance (MR) examinations, including diffusion-weighted imaging (DWI) and susceptibility weighted imaging (SWI), were included. Repeat MR examinations were performed within two weeks after admission. Patients were divided into hemorrhagic transformation (HT) group and non-HT group. The minimum and mean ADC values of the cerebral infarct areas were measured on the initial DWI. Between-group differences in ADC values were assessed using two-sample t-test. Results: Twenty-eight (48.3%) patients had developed HT, while no evidence of HT was observed in 30 (51.7%) patients. Patients with HT had lower minimum and mean ADC values (P < 0.05 vs. non-HT group). The optimal cut-off level of minimum ADC value for predicting HT was 450 × 10-6 mm2/s, and that of mean ADC value was 500 × 10-6 mm2/s. Conclusion: Change in signal intensity on the ADC map and quantitative analysis of the ADC values may help predict hemorrhagic transformation in patients with early ischemic infarction.

Keywords: Cerebral infarction, hemorrhagic transformation, diffusion-weighted imaging, apparent diffusion coef-ficient

Introduction

Ischemic cerebral infarction (also referred to as ischemic stroke) accounts for approximately 60%-80% of all cerebrovascular events [1]. Is- chemic stroke is more common than hemor-rhagic stroke and is a major contributor to mor-tality and long-term disability throughout the world [1, 2].

Hemorrhagic transformation (HT), which refers to secondary bleeding into the ischemic cere-bral area after stroke, is a serious complication [2, 3]. It may occur following anticoagulant and thrombolytic therapy as well [4]. HT is believed to be a consequence of ischemic injury to brain microvasculature. This complication occurs in 8.5%-30% of all patients with ischemic stroke [5]. Severe HT can worsen the patient’s condi-tion and may even be fatal. Since the treatment and prognosis of HT are distinct from those of

ischemic cerebral infarction, early prediction and identification of HT is a clinical imperative. Several clinical parameters and blood biomark-ers associated with an increased risk of HT have recently been reported [3, 6-8].

Magnetic resonance imaging (MRI) is widely used for the diagnosis and evaluation of cere-bral ischemia. Traditional T1-weighted imaging (T1WI) and T2-weighted imaging (T2WI) do not adequately delineate the acute lesions in the initial hours after ischemic stroke; moreover, these are not useful to assess the risk of HT. Diffusion-weighted imaging (DWI) is sensitive to the changes in the diffusion of water mole-cules associated with cytotoxic edema; the movement of these molecules is liable to be restricted in ischemic areas of the brain. The degree of this restriction can be quantified by measuring apparent diffusion coefficient (ADC) value [9-11]. DWI has been shown to be effec-

ADC value for prediction of hemorrhagic transformation

110 Int J Clin Exp Med 2018;11(1):109-117

tive in identification of early ischemic lesions. In recent studies, DWI was shown to delineate acute ischemic infarcts developing within 2-3 hours of ischemic stroke [12]. Indeed, use of DWI parameters to assess the risk of HT is an active area of research [3, 6-8]; however, defini-tive evidence of its clinical relevance is yet to be obtained. In this study, we investigated the role of ADC values measured on DWI for pre- diction of HT in patients with acute ischemic infarction.

ria: (1) no hemorrhagic focus detected on the initial susceptibility weighted imaging (SWI); (2) hypointense cerebral infarcts detected on fol-low-up SWI. Patients were divided into two groups (HT group and non-HT group) according to the presence or absence of HT.

Neuroimaging examination

MRI scan was conducted on a 1.5 T scanner (Signa HD; GE Medical Systems, Milwaukee,

Table 1. Magnetic resonance scanning parametersMagnetic reso-nance sequence

Repetition time/echo time

Section thickness/in-tersection gap (cm)

Bandwidth (Hz) NEX Matrix FOV Scan time

(minutes)T1WI 1500/24 6/2 27.78 2 320 × 224 24 × 24 2:17T2WI 4600/121 6/2 31 1.5 448 × 448 24 × 24 1:55DWI 6000/93.3 6/2 62.5 2 128 × 128 24 × 24 0:48SWI Min/54.1 2/0 41.67 0.69 448 × 320 24 × 24 4:00NEX, number excitations; FOV, field of view; T1WI, T1 weighted image; T2WI, T2 weighted image; DWI, diffusion weighted imag-ing; SWI, susceptibility weighting imaging; Min, minimum.

Figure 1. Magnetic resonance imaging of a 74-year-old man with a 5-hour-long history of glossolalia and left-limb weakness, but with no signs of hemor-rhagic transformation. A. Initial DWI showed cerebral infarction (arrowhead) in the left frontotemporal lobe. B. The initial ADC map showed hypointensity (arrowhead) in the infarcts; minimum ADC was 540 × 10-6 mm2/s and mean ADC was 640 × 10-6 mm2/s. C. Initial SWI demonstrated no hemorrhage in the infarction area. D. Follow-up SWI two weeks after admission showed no hemorrhagic transformation.

Materials and methods

Subjects

This prospective study was approved by the local ethics committee. In this study, we enrolled 58 patients with isch-emic cerebral infarction, who presented within 3 days from the onset of ischemia. Of th- ese, 11 patients had present-ed ≤ 6 hours since the onset of acute ischemic stroke. Wri- tten informed consent was obtained from all individual participants and/or their rela- tives.

The inclusion criteria were: (1) confirmed diagnosis of acute ischemic cerebral infarction; and (2) availability of complete clinical and radiological data. Exclusion criteria included: (1) findings suggestive of HT on the initial MRI; and (2) critical-ly ill patients, such as those with unstable blood pressure or coagulation disorders.

The diagnosis of HT was bas- ed on the following two crite-

ADC value for prediction of hemorrhagic transformation

111 Int J Clin Exp Med 2018;11(1):109-117

WI). The imaging protocol included T1WI, T2WI, DWI, and SWI sequences (all axial scans). MRI examination was performed on admission and two weeks thereafter. The imaging parameters are presented in Table 1. DWI was acquired with a diffusion sensitizing gradient (b values) of 0 and 1000 s/mm2.

Data processing and analysis

All DWI images were processed with the spe-cialized software, Functool of GE ADW 4.2 workshop. ADC analysis was performed on axial images, and the following sections were selected for measurement of ADC values. For evaluation on the initial MRI, in the HT group, we chose the section which corresponded with the follow-up MRI section that showed the larg-est area of the hemorrhagic lesion; in the non-HT group, we selected the section with the larg-est area of the infarct. The lateral cerebral

distribution of minimum and mean ADC values of the initial scans between the HT and the non-HT groups. Receiver operating characteristic (ROC) curve analysis was performed to assess the potential correlation of minimum and mean ADC values with the risk of HT. Log-binomial regression was used to estimate the relative risk of dichotomized minimum ADC and mean ADC values (cut-off values were set as 450 × 10-6 mm2/s and 500 × 10-6 mm2/s, respective-ly) on HT. A two-tailed P value < 0.05 was con-sidered indicative of a statistically significant difference.

Results

Clinical characteristics and HT analysis

There were 40 men and 18 women (age range, 35-89 years). The presenting clinical sympto- ms included dizziness, weakness of limbs, and

Figure 2. Magnetic resonance imaging of a 61-year-old man with a 5-hour-long history of glossolalia and left-limb weakness and with hemorrhagic transformation. A. Initial diffusion-weighted imaging (DWI) showed cerebral infarction (arrowhead) in the right peri-Rolandic areas. B. The initial appar-ent diffusion coefficient (ADC) map showed hypointensity (arrowhead) in the infarct; minimum ADC was 310 × 10-6 mm2/s and mean ADC was 390 × 10-6 mm2/s. C. Initial susceptibility weighted imaging (SWI) demonstrated no signs of hemorrhage in the infarct. D. Follow-up SWI two weeks after admis-sion showed patchy hypointensity (arrow) in the right peri-Rolandic areas, which indicated hemorrhagic transformation.

ventricle was manually excl- uded to minimize CSF con- tamination. Regions of inter-est (ROIs) were drawn on the T2W images, and automati-cally transferred onto the cor-responding ADC maps. Within each ROI, ADC values were determined on a pixel-by-pixel basis. The ADC values in the center of the infarct and at the four edges (top, bottom, left, and right) were measured, and the mean ADC value was cal- culated.

The SWI data was processed via rebuilding the minimum intensity projection. HT was de- fined as appearance of new hypointense lesions in the in- farcts on follow-up scans.

Statistical analysis

Statistical analysis was perfo- rmed using SAS software (ver-sion 9.4, SAS Institute, Cary, North Carolina). Two-sample t-test was used to compare the continuous variables; Chi-squared or Fisher exact test was used to compare the cat-egorical variables. Student t test was used to compare the

ADC value for prediction of hemorrhagic transformation

112 Int J Clin Exp Med 2018;11(1):109-117

slurring of speech. There was no evidence of HT (Figure 1) in 30 patients, out of which 6 were admitted within 6 hours from the onset of ischemia. On follow-up radiological examina-tion, 28 patients were found to have developed HT (Figures 2 and 3); five of these were admit-ted within 6 hours from the onset of ischemia.

(95% CI, 67.3%-96.0%) sensitivity and 93.3% (95% CI, 77.9%-99.2%) specificity. A cut-off value of 500 × 10-6 mm2/s for mean ADC value was associated with a sensitivity and specificity of 87.9% (95% CI, 74.5%-97.8%) and 95.6% (95% CI, 78.4%-99.5%), respectively (Figures 6 and 7).

Figure 3. Magnetic resonance imaging of a 75-year-old man with an half-an-hour-long history of left-limb weakness and with hemorrhagic transforma-tion. A. Initial DWI showed cerebral infarction (arrowhead) in the right frontal lobe and basal ganglia. B. The initial ADC map showed hypointensity (arrow-head) in the infarct; minimum ADC was 370 × 10-6 mm2/s and mean ADC was 440 × 10-6 mm2/s. C. Initial SWI demonstrated no hemorrhage in the infarct. D. Follow-up SWI two weeks after admission showed patchy hypoin-tensity (arrow) in the right frontal lobe and basal ganglia, which indicated hemorrhagic transformation.

No significant between-group differences were observed wi- th respect to general charac-teristics (Table 2).

DWI analysis: the minimum and mean ADC values

In the HT group, the minimum ADC values ranged between 250 × 10 -6 mm2/s and 470 × 10-6 mm2/s, while the mean ADC values ranged between 350 × 10-6 mm2/s and 530 × 10-6 mm2/s. In the non-HT gro- up, the minimum ADC values ranged between 390 × 10-6 mm2/s and 740 × 10-6 mm2/s, while the mean ADC values ranged between 460 × 10-6 mm2/s and 790 × 10-6 mm2/s (Figures 4 and 5). The mini-mum and mean ADC values in patients with HT were signifi-cantly lower than those in the non-HT group (two-sample t- test; P < 0.05) (Table 3).

ROC analysis; Optimal cut-off level of minimum ADC value and mean ADC value

ROC analysis revealed a high sensitivity and specificity of both minimum and mean ADC values in predicting HT after ischemic cerebral infarction. Area under the curve for mi- nimum and mean ADC was 97.3% (95% Confidence Inte- rval [CI], 93.6%-100%) and 98.1% (95% CI, 94.7%-100%), respectively.

A cut-off value of 450 × 10 -6 mm2/s for minimum ADC value was associated with 85.7%

Table 2. Demographic and clinical characteristics

CharacteristicsHT group Non-HT group

P valuen = 28 n = 30

Age (years; mean ± SD) 63.8 ± 10.2 65.7 ± 9.8 0.459Male [n (%)] 19 (67.9%) 21 (70.0%) 0.860Hypertension [n (%)] 23 (82.1%) 24 (80.0%) 0.835Diabetes [n (%)] 21 (75.0%) 23 (76.7%) 0.882Smoking [n (%)] 14 (50.0%) 11 (36.7%) 0.305Alcohol use [n (%)] 11 (39.3%) 7 (23.3%) 0.189HT, hemorrhagic transformation; SD, standard deviation.

ADC value for prediction of hemorrhagic transformation

113 Int J Clin Exp Med 2018;11(1):109-117

Discussion

Pathogenesis and diagnosis

The pathogenesis of HT after acute ischemic stroke is not completely understood. The main hypotheses include: (1) increased vascular per-meability following ischemic injury; (2) dislodge-ment of emboli and recanalization of occluded vessels; (3) high permeability of newly formed

microvasculature; and (4) reperfusion injury to small vessels, damage to blood brain barrier, and increased microvascular permeability after subsidence of ischemic edema [13, 14]. All these factors may contribute to bleeding.

DWI has been widely used as a gold standard for diagnosis of ischemic stroke. DWI is more sensitive than ordinary T2-weighted imaging for detection of ischemia within several hours of

Figure 4. Histogram showing the distribution of minimum ADC values.

Figure 5. Histogram showing the distribution of mean ADC values.

ADC value for prediction of hemorrhagic transformation

114 Int J Clin Exp Med 2018;11(1):109-117

the primary event [9-11]. Acute cerebral infarc-tion is associated with decrease in ADC values, which is attributable to the early pathophysio-logical changes including energy failure, mic- rocirculation disturbance, tissue acidosis and

cytotoxic edema [15]. In the early phase of isch-emia, cerebral hypoxia and metabolic derange-ment may cause swelling of neuronal cells, nar-rowing of extracellular space, and restricted diffusion of water molecules outside of cellular boundaries/membranes. Reduced overall dif-fusion level in ischemic tissues manifests as decreased ADC values and hyperintensity on DWI. However, DWI is not able to reflect HT in the early stages. SWI is considered as an effec-tive diagnostic modality for identification of HT owing to its exquisite sensitivity to venous bl- ood, hemorrhage and blood degradation prod-ucts. SWI is known to be the most sensitive MR sequence to detect different types of intracra-nial hemorrhage including microhemorrhages [16]. This imaging modality is based on differ-ences in magnetic susceptibility of different substances in a magnetic field. Deoxygenated hemoglobin, methemoglobin, and hemosiderin have strong paramagnetic properties that can cause local magnetic field inhomogeneity. He- morrhage causes a drop in signal (hypointensi-ty) on SWI. In addition, SWI was recently shown to detect post-ischemic cerebral hemorrhage within two hours after onset, which indicates its ability to detect HT in the hyperacute phase [17].

Application of ADC values

In recent studies, ADC values were shown to help predict the risk of HT in patients with isch-emic stroke [17, 18]. However, the precise na- ture of the association between ADC values and incidence of HT in these patients is not completely understood. There is a negative cor-relation between the ADC values and the DWI signals; restricted diffusion of water molecules lowers the ADC values, which manifests as hypointense lesions in the ADC map and hyper-intense lesions on DWI.

ADC value is known to reach its nadir within 8-32 hours and to maintain this low value for another 72 hours; MRI examinations should be done within this 72-hour time window. Based on this premise, the study assumes that the

Table 3. ADC values by study groupHT group n = 28 Non-HT group n = 30 95% Confidence interval P value

Minimum ADC value (mean ± SD) 384.3 ± 62.3 551.7 ± 73.9 131.5-203.3 < 0.001Mean ADC value (mean ± SD) 453.9 ± 43.8 620.0 ± 73.0 134.5-197.6 < 0.001HT, hemorrhagic transformation; ADC, apparent diffusion coefficient.

Figure 6. Receiver operating characteristic curve of minimum ADC value for prediction of HT.

Figure 7. Receiver operating characteristic curve of mean ADC value for prediction of HT.

ADC value for prediction of hemorrhagic transformation

115 Int J Clin Exp Med 2018;11(1):109-117

ADC values in the acute phase of cerebral infarction are relatively stable and thus a quan-titative comparison can be valid. Additionally, the peak incidence of HT is within 1 to 2 weeks after ischemia [18]. Therefore, the follow-up MR examinations for detecting HT were per-formed within 2 weeks after admission.

In the current study, statistical analysis showed significantly lower minimum and mean ADC val-ues in the HT group as compared to those in the non-HT group. We believe that the ADC value is expected to be a strong predictor for HT following acute ischemic cerebral infarction.

Studies have sought to determine the ADC threshold compatible with survival of ischemic tissues. Dardzinski et al. studied a rat model of ischemia, and suggested that the ADC thre- shold of permanent cerebral ischemia in two hours is 550 × 10-6 mm2/s [19]. Kazemi et al. blocked the middle cerebral artery for three hours, and found the absolute ADC values associated with necrosis to be < 470 × 10 -6 mm2/s [20]. Fiehler et al. studied the frequency of normalization of ADC values that were de- creased in hyperacute stroke, and noted that approximately 70%-80% of ADC value is recov-erable [21].

Clinical studies and animal experiments have confirmed that the severity of cerebral infarc-tion as well as the occurrence of HT has a direct relationship with the degree and duration of cerebral ischemia. Furthermore, the extent of damage to microvasculature and tight junc-tions between endothelial cells in the initial 24 hours also reflects the degree of cerebral isch-emia. In animal models, ischemic tissues with the minimum ADC values showed the most severe pathophysiological changes and dam-age to blood brain barrier, which corresponds to a high risk of HT [22-25]. Experiments have also suggested a correlation between decreas- ed ADC values in infarcts (including, minimum and mean ADC values) and HT. A more obvious decrease in ADC values usually indicates a higher risk of bleeding. Moreover, the bleeding area within the infarct mainly corresponds to the area with the lowest ADC value, and is mostly localized in the central area of the infarct; this further illustrates that the central part is most vulnerable to tissue necrosis.

Scholars have used different methods to deter-mine the cut-off values for prediction of HT.

Tong et al. measured the ADC values pixel-by-pixel in 17 patients with cerebral infarction within 8 hours of onset. They found that the ADC values in patients with HT were lower than those in patients with no evidence of HT, and that the pixels with ADC value < 550 × 10 -6 mm2/s in the HT group (47%) were significantly greater than those in the non-HT group (19%) [26, 27]. In addition, > 40% of the pixels showed ADC values < 550 × 10-6 mm2/s in all infarcts that eventually developed secondary HT. Never- theless, < 31% of the pixels had ADC values < 550 × 10-6 mm2/s in all non-HT lesions. Derek et al. recommended an ADC threshold of < 400 × 10-6 mm2/s for patients with cerebral infarc-tion who underwent intravenous rt-PA throm-bolysis within 7 hours of onset; the pixels of ADC value < 400 × 10-6 mm2/s in the HT group were significantly higher than those in the non-HT group [28].

In the current study, the cut-off level of mini-mum ADC values for predicting HT was 450 × 10-6 mm2/s, and that of mean ADC values was 500 × 10-6 mm2/s. The difference between our findings and those reported previously may be due to the following reasons: (1) different time-points of MRI examination: the patients in this study underwent MRI scanning within 3 days after onset of ischemia as many patients re- ached hospital > 6 hours after onset of symp-toms. The time window was only 3-8 hours in the earlier studies; (2) different treatment: pa- tients in this study did not receive any thrombo-lytic therapy with the exception of anti-platelet therapy, while patients in the earlier studies underwent intravenous thrombolysis.

Conclusion

This study indicates that DWI is a useful tool to evaluate acute ischemic infarction, and that the quantitative analysis of ADC values is help-ful in the assessment. SWI is a sensitive meth-od to identify secondary HT. The ADC values in infarcts can reflect the degree of ischemia and may be a useful reference for predicting the risk of HT in patients with early ischemic in- farction.

Disclosure of conflict of interest

None.

Address correspondence to: Xin-Qing Jiang, The First Affiliated Hospital of Jinan University, West

ADC value for prediction of hemorrhagic transformation

116 Int J Clin Exp Med 2018;11(1):109-117

Whampoa Avenue No. 613, Guangzhou 510630, China; Guangzhou First People’s Hospital, Panfu Road No. 1, Guangzhou Medical University, Guang- zhou 510180, China. Tel: 18102812838; E-mail: 540639601@qq.com; Xiu-He Wang, The First Affi- liated Hospital of Jinan University, West Whampoa Avenue No. 613, Guangzhou 510630, China. Tel: 13610353278; E-mail: twxh@jnu.edu.cn

References

[1] Giossi A, Volonghi I, Del Zotto E, Costa P, Pado-vani A and Pezzini A. Large middle cerebral ar-tery and panhemispheric infarction. Front Neu-rol Neurosci 2012; 30: 154-157.

[2] Yen P, Cobb A and Shankar JJ. Does computed tomography permeability predict hemorrhagic transformation after ischemic stroke? World J Radiol 2016; 8: 594-599.

[3] Liu J, Wang D, Li J, Xiong Y, Liu B, Wei C, Wu S, Lin J and Liu M. Increased serum alkaline phosphatase as a predictor of symptomatic hemorrhagic transformation in ischemic stroke patients with atrial fibrillation and/or rheumat-ic heart disease. J Stroke Cerebrovasc Dis 2016; 25: 2448-2452.

[4] Tan Z, Li X, Turner RC, Logsdon AF, Lucke-Wold B, DiPasquale K, Jeong SS, Chen R, Huber JD and Rosen CL. Combination treatment of r-tPA and an optimized human apyrase reduces mortality rate and hemorrhagic transformation 6h after ischemic stroke in aged female rats. Eur J Pharmacol 2014; 738: 368-373.

[5] Paciaroni M, Agnelli G, Corea F, Ageno W, Al-berti A, Lanari A, Caso V, Micheli S, Bertolani L, Venti M, Palmerini F, Biagini S, Comi G, Previdi P and Silvestrelli G. Early hemorrhagic trans-formation of brain infarction: rate, predictive factors, and influence on clinical outcome: re-sults of a prospective multicenter study. Stroke 2008; 39: 2249-2256.

[6] Yang N, Lin M, Wang BG, Zeng WY, He YF, Peng HY, Zeng J, Wu ZY and Zhong Y. Low level of low-density lipoprotein cholesterol is related with increased hemorrhagic transformation af-ter acute ischemic cerebral infarction. Eur Rev Med Pharmacol Sci 2016; 20: 673-678.

[7] Fagan SC, Lapchak PA, Liebeskind DS, Ishrat T and Ergul A. Recommendations for preclinical research in hemorrhagic transformation. Transl Stroke Res 2013; 4: 322-327.

[8] Chen X, Wang Y, Fu M, Lei H, Cheng Q and Zhang X. Plasma immunoproteasome predicts early hemorrhagic transformation in acute is- chemic stroke patients. J Stroke Cerebrovasc Dis 2016; 26: 49-56.

[9] Perez-Trepichio AD, Xue M, Ng TC, Majors AW, Furlan AJ, Awad IA and Jones SC. Sensitivity of

magnetic resonance diffusion-weighted imag-ing and regional relationship between the ap-parent diffusion coefficient and cerebral blood flow in rat focal cerebral ischemia. Stroke 1995; 26: 667-674; discussion 674-665.

[10] Minematsu K, Li L, Fisher M, Sotak CH, Davis MA and Fiandaca MS. Diffusion-weighted mag-netic resonance imaging: rapid and quantita-tive detection of focal brain ischemia. Neurol-ogy 1992; 42: 235-240.

[11] Kate MP, Riaz P, Gioia L, Sivakumar L, Jeer-akathil T, Buck B, Beaulieu C and Butcher K. Dynamic evolution of diffusion-weighted im- aging lesions in patients with minor ischemic stroke. Stroke 2015; 46: 2318-2321.

[12] Hanyu H, Imon Y, Sakurai H, Iwamoto T, Taka-saki M, Shindo H and Abe K. Diffusion-weight-ed magnetic resonance and magnetization transfer imaging in the assessment of isch-emic human stroke. Intern Med 1998; 37: 360-365.

[13] Marsh EB, Llinas RH, Schneider AL, Hillis AE, Lawrence E, Dziedzic P and Gottesman RF. Predicting hemorrhagic transformation of ac- ute ischemic stroke: prospective validation of the HeRS score. Medicine (Baltimore) 2016; 95: e2430.

[14] D’Amelio M, Terruso V, Famoso G, Di Benedet-to N, Realmuto S, Valentino F, Ragonese P, Sa-vettieri G and Aridon P. Early and late mortality of spontaneous hemorrhagic transformation of ischemic stroke. J Stroke Cerebrovasc Dis 2014; 23: 649-654.

[15] Gersing AS, Ankenbrank M, Schwaiger BJ, Toth V, Janssen I, Kooijman H, Wunderlich S, Bauer JS, Zimmer C and Preibisch C. Mapping of ce-rebral metabolic rate of oxygen using dynamic susceptibility contrast and blood oxygen level dependent MR imaging in acute ischemic stroke. Neuroradiology 2015; 57: 1253-1261.

[16] Haacke EM, Mittal S, Wu Z, Neelavalli J and Cheng YC. Susceptibility-weighted imaging: technical aspects and clinical applications, part 1. AJNR Am J Neuroradiol 2009; 30: 19-30.

[17] Liu C, Dong Z, Xu L, Khursheed A, Dong L, Liu Z, Yang J and Liu J. MR image features predict-ing hemorrhagic transformation in acute cere-bral infarction: a multimodal study. Neuroradi-ology 2015; 57: 1145-1152.

[18] Adami A, Thijs V, Tong DC, Beaulieu C, Moseley ME and Yenari MA. Use of diffusion weighted MRI to predict the occurrence and severity of hemorrhagic transformation in a rabbit model of embolic stroke. Brain Res 2002; 944: 32-39.

[19] Dardzinski BJ, Sotak CH, Fisher M, Hasegawa Y, Li L and Minematsu K. Apparent diffusion coefficient mapping of experimental focal cere-

ADC value for prediction of hemorrhagic transformation

117 Int J Clin Exp Med 2018;11(1):109-117

bral ischemia using diffusion-weighted echo-planar imaging. Magn Reson Med 1993; 30: 318-325.

[20] Kazemi M, Silva MD, Li F, Fisher M and Sotak CH. Investigation of techniques to quantify in vivo lesion volume based on comparison of wa-ter apparent diffusion coefficient (ADC) maps with histology in focal cerebral ischemia of rats. Magn Reson Imaging 2004; 22: 653-659.

[21] Fiehler J, Knudsen K, Kucinski T, Kidwell CS, Alger JR, Thomalla G, Eckert B, Wittkugel O, Weiller C, Zeumer H and Rother J. Predictors of apparent diffusion coefficient normalization in stroke patients. Stroke 2004; 35: 514-519.

[22] del Zoppo GJ, von Kummer R and Hamann GF. Ischaemic damage of brain microvessels: in-herent risks for thrombolytic treatment in stroke. J Neurol Neurosurg Psychiatry 1998; 65: 1-9.

[23] Knight RA, Dereski MO, Helpern JA, Ordidge RJ and Chopp M. Magnetic resonance imaging assessment of evolving focal cerebral isch-emia. Comparison with histopathology in ra- ts. Stroke 1994; 25: 1252-1261; discussion 1261-1252.

[24] Miyasaka N, Nagaoka T, Kuroiwa T, Akimoto H, Haku T, Kubota T and Aso T. Histopathologic correlates of temporal diffusion changes in a rat model of cerebral hypoxia/ischemia. AJNR Am J Neuroradiol 2000; 21: 60-66.

[25] Dijkhuizen RM, Asahi M, Wu O, Rosen BR and Lo EH. Delayed rt-PA treatment in a rat embolic stroke model: diagnosis and prognosis of isch-emic injury and hemorrhagic transformation with magnetic resonance imaging. J Cereb Blood Flow Metab 2001; 21: 964-971.

[26] Tong DC, Adami A, Moseley ME and Marks MP. Prediction of hemorrhagic transformation fol-lowing acute stroke: role of diffusion- and per-fusion-weighted magnetic resonance imaging. Arch Neurol 2001; 58: 587-593.

[27] Tong DC, Adami A, Moseley ME and Marks MP. Relationship between apparent diffusion coef-ficient and subsequent hemorrhagic transfor-mation following acute ischemic stroke. Stroke 2000; 31: 2378-2384.

[28] Derex L, Hermier M, Adeleine P, Pialat JB, Wiart M, Berthezene Y, Philippeau F, Honnorat J, Fro-ment JC, Trouillas P and Nighoghossian N. Clinical and imaging predictors of intracerebral haemorrhage in stroke patients treated with intravenous tissue plasminogen activator. J Neurol Neurosurg Psychiatry 2005; 76: 70-75.