Introduction

Ischemic injury is heterogeneous, including infarct core and ischemic penumbra，which may partially recover with prompt treatment1. DWI is a reliable neuroimaging technique for ischemic stroke assessment. DWI model approximates biological water diffusion as being Gaussian diffusion. However, non-Gaussian diffusion is often observed throughout the brain, including ischemic tissue damage, tumor and so on2. Thus, information obtained using conventional DWI may be incomplete for ischemic injury. Diffusion kurtosis imaging is an emerging MRI technique that measures the degree of the non-Gaussian water diffusion, which offers information complimentary to conventional diffusion metrics and is sensitive to detecte microstructural cerebral tissue changes in stroke3. The study has demonstrated that the MD-DKI mismatch was highly correlated with the three-month infarct size on T2WI image4. Moreover, MD-DKI mismatch can further grade the acute ischemic tissue status for ischemic damage, which enhances the ability to guide the treatment of ischemic tissue. The purpose of this study is to explore a more sensitive biomarker to predict the clinical outcome in focal ischemic stroke at day 90, and to observe the correlation between clinical outcome and MD-DKI metrics mismatch whether treated with or not with intravenous tPA.

Methods

We prospectively recruited 58 patients (age 56 ±7.8 years old) with acute neurological deficit compatible with focal ischemic stroke, who also studied with DKI within 24 hours of symptom onset. The study was approved by local IRB. Patients with a history of prior significant stroke, and other neurological deficit or intracranial hemorrhages were excluded from study. The patients were randomly divided into four groups on basis of MD-DKI match or mismatch and treatment with or without intravenous tPA. MRI and the National Institutes of Health Stroke Scale（NIHSS） at day 1, day 90 after symptom onset were analyzed. MRI was performed on a 3.0T scanner (ingenia, Philips Medical Systems,The Netherlands), including T1WI, T2WI, DWI and DKI. DKI sequences were acquired with 3 b-values (0,1000 and 2000s/mm2) along 30 diffusion encoding directions with following parameters: TR/TE:2440/83ms, FOV=224x224mm2, acquisition matrix=112×112,image resolution=2.5×2.5x2.5mm2, acceleration factor=3, slice thickness=5 mm (no gap), acquisition time approximately 15 minutes. Image analysis: The DKI data, motion and eddy current correction were processed with FMRIB Software Library (FSL, Oxford, United Kingdom). The post process was performed on DKI Toolbox software in MATLAB, Parametric maps for MK, RK, AK and MD were subsequently obtained. ROIs were manually identified by the hyperintensity on the diffusion-weighted images, ROIs were included the ischemic lesion with pixel values distinctly higher than the contralateral hemisphere, and any ambiguous pixels were not included. Percent changes（（ROIMD-ROI DKI-metrics）/ROIMD）of lesion volume on DWI and DKI metrics from to ischemic tissue were computed. We performed a one-way analysis of variance (ANOVA) among the contralateral normal group, match group and mismatch group. The student t-test was used to compare DKI metrics, MD values and percent change of lesion volume in lesional tissue with or without intravenous tPA treatment. We performed a Pearson correlation to test for the possible correlations between changes in the DKI metrics and percent change at day 1 and 90. P<0.05 is considered statistically significant

Results

32 (55.1%) focal ischemic stroke patients had DWI–DKI mismatch. MK,AK and RK were significantly higher in lesions compared with contralateral normal group at day 1 after stroke(Fig.1). MD was lower in lesions than contralateral normal at days 90. MK and AK of lesions in mismatch groups were lower than match groups (p<0.05) (Table.1). Intravenous tPA was significantly associated with good clinical outcome inmismatch group compared with no thrombolysis (p=0.01) (Table.2). We found that the percent changes (△MD-AK) in the ischemic lesion volume at day 1 after stroke were correlated with the percent changes (△NIHSS) of NIHSS at day 1 and day 90(r=0.721,p=0.002)(Fig.2).

Discussion

This study confirmed that DKI indices were sensitive to pathological changes and microstructural complexity in focal ischemic stroke. Our results indicated that MK and AK of lesions in mismatch groups were lower than match groups and MD were higher than match groups, These results proved to be dominated by the change in the intra-axonal microenvironment of early ischemic damage tisseue. MD-AK mismatch recovered reasonably well with intravenous tPA at onset of stroke, whereas regions with coincidence MD-AK showed poor recovery. Thrombolysis based on MD-AK mismatch yielded a substantially better clinical outcome at day 90 after stroke. MD-AK mismatch might reflect the simultaneous presence the infarct core and the ischemic penumbra in lesions, which was similar to perfusion–diffusion mismatch. Interpretation MD-AK mismatch could be used to identify patients from baseline imaging in focal ischemic stroke patients who were likely to benefit from intravenous thrombolysis. AK might be an effective biomarker to accurately define the range of infarct core and ischemic penumbra.

Conclusion

Diffusion kurtosis imaging can provide the complexity and heterogeneity for enhanced characterization of ischemic tissue injury. We demonstrated that MD-AK mismatch could be used to identify focal ischemic stroke patients from baseline DKI who are likely to benefit from intravenous thrombolysis at onset of stroke.

Acknowledgements

The authors would like to thank Philips Healthcare for their technical assistance.