Sage Journals: Discover world-class research

Abstract

Diffusion weighted imaging (DWI) has been commonly used in acute stroke examination, yet a portion of DWI lesion may be salvageable. Recently, it has been shown that diffusion kurtosis imaging (DKI) defines the most severely damaged DWI lesion that does not renormalize following early reperfusion. We postulated that the diffusion and kurtosis lesion mismatch experience heterogeneous hemodynamic and/or metabolic injury. We investigated tissue perfusion, pH, diffusion, kurtosis and relaxation from regions of the contralateral normal area, diffusion lesion, kurtosis lesion and their mismatch in an animal model of acute stroke. Our study revealed significant kurtosis and diffusion lesion volume mismatch (19.7 ± 10.7%, P < 0.01). Although there was no significant difference in perfusion and diffusion between the kurtosis lesion and kurtosis/diffusion lesion mismatch, we showed lower pH in the kurtosis lesion (pH = 6.64 ± 0.12) from that of the kurtosis/diffusion lesion mismatch (6.84 ± 0.11, P < 0.05). Moreover, pH in the kurtosis lesion and kurtosis/diffusion mismatch agreed well with literature values for regions of ischemic core and penumbra, respectively. Our work documented initial evidence that DKI may reveal the heterogeneous metabolic derangement within the commonly used DWI lesion.

Keywords

Acute stroke amide proton transfer chemical exchange saturation transfer (CEST)diffusion kurtosis imaging (DKI)diffusion weighted imaging (DWI)pH

Introduction

Diffusion weighted imaging (DWI) is sensitive to acute stroke and has been widely regarded as an imaging approximation of the most severely damaged infarction core.^1,2 Built on this assumption, multiple mismatch paradigms have been proposed that use perfusion/DWI, angiography/DWI, fluid-attenuated inversion recovery imaging (FLAIR)/DWI, or clinical/DWI mismatches to identify potentially salvageable ischemic tissue (penumbra).^3–7 In addition, infarction core volume, often assessed by diffusion MRI, has been used as one of imaging criteria to guide late stroke treatment.^8–10 However, increasing evidence has suggested that DWI lesion includes both the infarction core and potentially salvageable ischemic tissue; a portion of DWI lesion may normalize following effective recanalization, even in cases of large DWI lesions.^11,12 This is consistent with the observations of graded metabolic disruption within the DWI lesion.^13,14 It is not uncommon to have sustained DWI reversibility, especially with early reperfusion and therefore, the use of DWI lesion as infarction core may underestimate ischemic penumbra volume.^15,16 However, the severity of apparent diffusion coefficient (ADC) change does not predict irreversible tissue damage and there has been no well-established imaging means to delineate the heterogeneous DWI lesion.¹⁷

Diffusion kurtosis imaging (DKI) quantifies non-Gaussian diffusion that refines the commonly used yet simplistic ADC analysis.^18–21 It has been shown that kurtosis/diffusion lesion mismatch often renormalizes following prompt recanalization while kurtosis lesion shows poor recovery, suggesting that kurtosis lesion identifies the most severely damaged DWI lesion.²² However, little is known about the underlying pathophysiological difference between kurtosis lesion and kurtosis/diffusion lesion mismatch; positron emission tomography (PET) is not readily available in the acute stroke setting and the spatiotemporal resolution of MR spectroscopy is limited, particularly in experimental stroke models. It is worthwhile to mention that tissue acidosis is a surrogate metabolic index that alters prior to tissue infarction.²³ Therefore, we postulated that DKI may reflect graded hemodynamic and/or metabolic injury within the heterogeneous DWI lesion.

In this study, we used a well-established middle cerebral artery occlusion (MCAO) rodent model of acute stroke, and collected multi-parametric MRI to assess ischemic tissue injury. Specifically, we acquired arterial spin labeling (ASL) MRI to measure tissue perfusion. To investigate pH, we applied amide proton transfer (APT) MRI, a specific form of chemical exchange saturation transfer (CEST) MRI that is sensitive to pH-dependent amide proton exchange between endogenous mobile proteins/peptides and bulk tissue water.^24–27 It has been shown that APT MRI is closely correlated with lactate concentration during acute ischemic stroke, and tissue pH can be estimated from quantitative APT MRI.^28,29 To suppress the intrinsic kurtosis heterogeneity not related to acute ischemia, we applied the inherent correlation-based normalization (ICON) analysis for semiautomatic kurtosis lesion segmentation.³⁰ We then analyzed perfusion, pH, diffusion, kurtosis, and relaxation values in the contralateral normal area, diffusion lesion, kurtosis lesion, and the kurtosis/diffusion lesion mismatch. Importantly, we found that kurtosis lesion had significantly worsened pH drop from the kurtosis/diffusion lesion mismatch, substantiating the postulation that DKI provides an expedient means to resolve the routine DWI lesion based on its severity of metabolic derangement.

Material and methods

Rodents

In vivo studies have been approved by the institutional animal care and use committee, Massachusetts General Hospital (IACUC, MGH). All procedures were conducted in accordance with the National Research Council guidelines for the care and use of animals. Our study is also in compliance with the ARRIVE guidelines. Briefly, adult male Wistar rats (Charles River Laboratory, Wilmington, MA) were anesthetized with 1.5–2.0% isoflurane/air mixture throughout the study, with physiological parameters (e.g. heart rate, sPO₂, and rectal temperature) monitored and maintained within their normal ranges. Five normal rats (n = 5) and additional fifteen acute stroke rats (n = 15) were imaged between 1 and 2 h after MCAO. Briefly, after exposure of the right carotid bifurcation, the common carotid and distal external carotid arteries were sutured, and a silicone-coated 4-0 nylon filament was inserted and slowly advanced to block the origin of middle cerebral artery. One animal had failed MCAO surgery and was removed from the study. Another animal had technical issues with ASL MRI, and perfusion scan was excluded from data analysis.

MRI

Images (five slices, slice thickness = 2 mm, field of view = 20 × 20 mm², matrix = 48 × 48) were acquired at a 4.7 T Biospec MRI (Bruker Biospin, Billerica, MA). Briefly, we scanned a fast DKI protocol (gradient duration/diffusion time (δ/Δ) = 6/20 ms, repetition time (TR)/echo time (TE) = 2500/36.6 ms, 4 averages, scan time = 3 min).³¹ The fast DKI protocol consisted of one b = 0 s/mm² reference image, followed by three images of b = 1000 s/mm² along three gradient directions of (1,0,0), (0,1,0), and (0,0,1), and nine images of b = 2500 s/mm² along nine diffusion directions of ${\hat{n}}^{(i)}$ , ${\hat{n}}^{(i +)}$ , and ${\hat{n}}^{(i -)}$ , which were defined as ${\hat{n}}^{(1)} = (1, 0, 0)^{T}$ , ${\hat{n}}^{(1 +)} = (0, 1, 1)^{T}$ , and ${\hat{n}}^{(1 -)} = (0, 1, - 1)^{T}$ , and similarly for i = 2 and 3. Note that the superscript i in ${\hat{n}}^{(i)}$ labels the position of the “1” while in ${\hat{n}}^{(i +)}$ and ${\hat{n}}^{(i -)}$ it labels the position of the “0.” T₁-weighted images were acquired using inversion recovery echo planar imaging (EPI), with seven inversion delays from 250 to 3000 ms (recovery time = 6500 ms, TE = 15 ms, 4 averages, scan time = 4 min); T₂-weigthed images were obtained with two TEs of 30 and 100 ms (TR = 3250 ms, 4 averages, scan time = 0.5 min). We collected amplitude-modulated ASL MRI (TR/TE =6500/15 ms, B₁ = 4.7 μT, time of saturation = 3250 ms, labeling distance of 15 mm, modulation frequency of 250 Hz, 32 averages and scan time = 7 min). pH MRI was acquired with fast APT MRI with unevenly segmented RF irradiation (recovery time = 3000 ms, primary RF saturation time = 3000 ms, secondary RF saturation time = 500 ms, and B₁ = 0.75 μT). The unsaturated control scan was signal-averaged 8 times, while the saturated images were averaged 32 times (scan time = 4 min).

Data analysis

MRI images were processed in Matlab (Mathworks, Framingham, MA). Specifically, for the fast DKI, mean diffusivity (MD) was calculated as the mean of MD_x,y,z derived from the formula proposed by Jensen et al.¹⁹:

{MD}_{x, y, z} = \frac{(b_{1} + b_{3}) D_{x, y, z}^{(12)} - (b_{1} + b_{2}) D_{x, y, z}^{(13)}}{b_{3} - b_{2}}

(1)

where

D_{x, y, z}^{(ij)} = \frac{{lnS (b}_{i} {) / S (0) - lnS (b}_{j}) / S (0)}{b_{j} - b_{i}}

, i = 1, j = 2, 3, and b₁ = 0, b₂ = 1000, and b₃ = 2500 s/mm². The mean kurtosis (MK) was obtained using the method described by Hansen et al.³¹:

M K = \frac{\frac{6}{15} (\sum_{i = 1}^{3} \ln \frac{S (b_{3}, {\hat{n}}^{(i)})}{S (0)} + 2 \sum_{i = 1}^{3} \ln \frac{S (b_{3}, {\hat{n}}^{(i +)})}{S (0)} + 2 \sum_{i = 1}^{3} \ln \frac{S (b_{3}, {\hat{n}}^{(i -)})}{S (0)}) + 6 \cdot b_{3} \cdot M D}{b_{3}^{2} \cdot M D^{2}}

(2)

In addition, parametric T₁ map was obtained with least squares fitting of the signal as a function of the inversion time ( $I ({TI}_{i}) = I_{0} [1 - (1 - η) e^{- {TI}_{i} / T_{1 w}}]$ ), where η is the inversion efficiency and TI_i is the ith inversion time. T_2w map was calculated as $T_{2 w} = \frac{{TE}_{2} - {TE}_{1}}{\ln (I ({TE}_{1}) / I ({TE}_{2}))}$ , where TE_1,2 are two TEs. In addition, cerebral blood flow (CBF) was derived as $CBF = \frac{λ (I_{ref} - I_{tag})}{2 α \cdot I_{ref}} \cdot \frac{e^{w / T_{1 a}}}{T_{1 w}}$ , where I_tag is the label image, I_ref is the reference image, λ is the brain–blood partition coefficient for water, α is the degree of inversion with transient time correction, w is the post-labeling delay, and T_1a is the arterial blood longitudinal relaxation time. ΔCBF map was calculated as the difference of pixel-wised CBF with the mean value of CBF in the contralateral normal brain. Tissue pH was derived from pH-weighted APT MRI using the calibrated base-catalyzed amide proton exchange relationship.²⁹

For kurtosis lesion segmentation, we generalized the ICON analysis by including both magnetization transfer (MT) and T₁ images.³⁰ Briefly, pixel-wised MK, T₁ and mean magnetization transfer ratio (MMTR) indices from the contralateral normal brain, excluding ventricles, was correlated using Pearson’s correlation. An estimated MK map (MK_est) was calculated per pixel (i.e. MK_est =C₀ + C₁*R₁ + C₂*MMTR + C₃*R₁*MMTR), where C_s are coefficients determined from the intact normal brain. The difference between the experimentally measured MK and that estimated from regression analysis assuming no ischemia was calculated ΔMRMK = MK-MK_est. The diffusion and kurtosis (i.e. ICON-MK) lesions were determined using a K-means clustering-based algorithm.³⁰ Note that we defined kurtosis lesion as kurtosis abnormality, similar as the commonly used perfusion and diffusion lesions. Contrast to noise ratio (CNR) between striatum and cortex regions was calculated as $CNR = | S_{striatum} - S_{cortex} | / \sqrt{(σ_{striatum}^{2} + σ_{cortex}^{2}) / 2}$ , where S and σ are the signal (i.e. MD, MK and ΔMRMK) and their standard deviations from the regions of striatum and cortex, respectively. We performed Kruskal–Wallis test with Dunns posttest analysis, and P value less than 0.05 was considered statistically significant.

Results

Multiparametric images from a representative normal rat are shown in Figure 1. Specifically, Figure 1(a) and (b) shows R₂ (1/T₂) and R₁ (1/T₁) images, respectively. In addition, ΔCBF (Figure 1(c)) map shows no noticeable hypoperfusion, as expected. MD map (Figure 1(d)) was reasonably homogeneous within the brain while MK image (Figure 1(e)) shows noticeable heterogeneity between the white matter (e.g. striatum) and gray matter (e.g. cortex, 0.79 ± 0.05 vs. 0.63 ± 0.04, P < 0.01). We assessed the image heterogeneity using the CNR between striatum and cortex regions. The CNR for MD and MK images was 0.57 ± 0.17 and 2.37 ± 0.49, respectively. The high CNR from the MK image suggested substantial intrinsic heterogeneity in the intact brain tissue. We found significant correlation between MK and R₁ (Figure 1(b)) and MMTR (Figure 1(f)), per pixel (Figure 1(g), R²= 0.60 ± 0.09, P < 0.01). Indeed, the difference between the experimentally measured MK image and that estimated assuming no ischemia insult (Figure 1(h), ΔMRMK =MK-MK_est) is substantially more homogeneous than the raw MK image. The ICON analysis significantly reduced the kurtosis CNR between striatum and cortex from 2.37 ± 0.49 (MK) to 0.39 ± 0.16 (P < 0.01), without a significant difference from that of MD (P > 0.10). Therefore, both MD and ΔMRMK images have CNR well under 1, suggesting that their regional signal variations in the intact brain tissue are within their respective noise level, rendering their change specific to ischemic tissue (Table 1).

Figure 1.

Illustration of magnetization transfer and relaxation-normalized kurtosis image. (a) R₂ image, (b) R₁ image, (c) ΔCBF image, (d) MD image, (e) experimentally measured MK image, (f) MMTR image, (g) regression between MK and MMTR and R₁, and (h) the ΔMRMK map that is of reduced intrinsic MK heterogeneity not related to acute ischemia.

Table 1.

Magnetization transfer and relaxation normalized MK map reduces intrinsic heterogeneity between intact brain WM and GM.

	Striatum (WM)	Cortex (GM)	CNR
MD (µm²/ms)	0.83 ± 0.01	0.81 ± 0.02	0.57 ± 0.17
MK	0.79 ± 0.05	0.63 ± 0.04	2.37 ± 0.49
ΔMRMK	0.05 ± 0.01	0.05 ± 0.01	0.39 ± 0.16

Figure 2 shows R₂ (Figure 2(a)) and R₁ (Figure 2(b)) maps from a representative acute stroke rat, displaying only small changes. Notably, ΔCBF (Figure 2(c)) and MD (Figure 2(d)) maps showed ipsilateral hypoperfusion and diffusion lesion, showing severe ischemia, with MK and ΔMRMK images shown in Figure 2(e) and (f), respectively. Whereas MD image presents significant diffusion lesion in the ischemic brain, the ipsilateral lateral cortex, interestingly, appeared to have relatively little kurtosis elevation when compared to that of the striatum. Both diffusion and kurtosis (ΔMRMK) lesions were semiautomatically segmented and overlaid on a DWI image (b = 1000 s/mm²), with kurtosis lesion shown in black, and kurtosis/diffusion lesion mismatch shown in pink (Figure 2(g)). Over all animals, the kurtosis lesion and diffusion lesion volume were 122.3 ± 47.4 mm³ and 153.4 ± 64.1 mm³, respectively (P < 0.01, paired t-test). The kurtosis/diffusion lesion mismatch was 19.7 ± 10.7% of the DWI lesion volume (P < 0.01, two-tailed one sample t-test). Importantly, tissue pH map (Figure 2(h)) revealed that the kurtosis/diffusion lesion mismatch had less pH drop when compared to that of the kurtosis lesion.

Figure 2.

Multiparametric MRI from a representative MCAO rat. (a) R₂ image, (b) R₁ image, (c) ΔCBF image, (d) MD image, (e) experimentally measured MK image, (f) ΔMRMK image, (g) kurtosis/diffusion lesion mismatch overlaid on a DWI image (b = 1000 s/mm²), and (h) pH image.

The diffusion, kurtosis, perfusion, pH, and relaxation indices were evaluated for the regions of the contralateral normal area, diffusion lesion, kurtosis lesion, and their mismatch, using Kruskal–Wallis test (Table 2). Although MD in the diffusion lesion, kurtosis lesion and kurtosis/diffusion lesion mismatch all reduced significantly from that of the normal region, there was no significant difference in MD among the lesions (Figure 3(a)). Figure 3(b) shows that ΔMRMK in the diffusion lesion and kurtosis lesion was significantly elevated from that of the normal area, and the kurtosis/diffusion mismatch has noticeably lower ΔMRMK from the kurtosis lesion. In addition, CBF only captured hypoperfusion in ischemic lesions from the reference area without significant difference among the lesions (Figure 3(c)), similar to MD image. Moreover, Figure 3(d) shows that pH in diffusion lesion (6.68 ± 0.11) and kurtosis lesion (6.64 ± 0.12) significantly decreased from the reference value (7.01 ±0.04, P < 0.05). Importantly, pH in the kurtosis/diffusion mismatch area (6.84 ± 0.11) was significantly higher than that of the kurtosis lesion (6.64 ± 0.12, P < 0.05). The diffusion lesion, kurtosis lesion, and kurtosis/diffusion lesion mismatch exhibited significantly higher T₁ than that of the normal area; meanwhile, only kurtosis lesion showed substantial T₂ augmentation compared to other regions.

Table 2.

Multi-parametric MRI indexes for the normal, diffusion lesion, kurtosis lesion, and their mismatch.

	Contralateral normal	MD lesion	MK lesion	MD/MK mismatch
MD (µm²/ms)	0.88 ± 0.03	0.60 ± 0.04*	0.60 ± 0.04*	0.67 ± 0.05*
ΔMRMK	0.05 ± 0.01	0.35 ± 0.04*	0.43 ± 0.04*	0.15 ± 0.03^†
ΔCBF (ml/g.min)	0.0	−0.66 ± 0.16*	−0.66 ± 0.20*	−0.52 ± 0.20*
pH	7.01 ± 0.04	6.68 ± 0.11*	6.64 ± 0.12*	6.84 ± 0.11^†
T₁ (s)	1.54 ± 0.05	1.63 ± 0.07*	1.61 ± 0.06*	1.64 ± 0.07*
T₂ (ms)	53.83 ± 0.99	52.94 ± 1.31	56.18 ± 1.20*^,‡	53.71 ± 1.48^†

Kruskal–Wallis test with Dunns posttest analyses were performed with *P < 0.05 indicating significant difference between contralateral normal and lesion regions, ^†P < 0.05 indicating significant difference between MK lesion and MD/MK lesion mismatch, ^‡P < 0.05 indicating significant difference between MD lesion and MK lesion or MD/MK lesion mismatch, respectively.

Figure 3.

Regional analysis of multi-parametric MRI of the normal region, diffusion lesion, kurtosis lesion, and kurtosis/diffusion lesion mismatch. (a) MD, (b) ΔMRMK, (c) ΔCBF, (d) pH, (e) T₁, and (f) T₂.

Discussion

Our study documented significant pH difference between the kurtosis lesion and kurtosis/diffusion lesion mismatch, providing converging evidence that the commonly used and heterogeneous DWI lesion can be delineated based on emerging pH and DKI stroke MRI. It is noteworthy to mention that Peek et al. used double autoradiographic imaging and reported that penumbral tissue has significantly elevated rate of glucose consumption and higher pH (6.87 ± 0.05) from the infarction core, which has lower pH (6.69 ± 0.11, P < 0.05) and reduced glucose metabolism.³² Our study found pH in the kurtosis lesion and kurtosis/diffusion mismatch in good agreement with those determined by Peek et al., suggesting that kurtosis lesion is more specific to infarction core while the kurtosis/diffusion mismatch belongs to penumbra.

It is important to point out that both perfusion and diffusion methods provide incomplete tissue classification; PWI lesion overestimates the outer boundary of penumbra by including benign oligemia, while the DWI lesion underestimates the inner boundary of penumbra as part of the DWI lesion is reversible.^15,33 As such, our study focused on refining the commonly used diffusion lesion, and additional study is needed to improve perfusion imaging for penumbral tissue stratification. Specifically, the commonly used ADC analysis assumes that water molecules follow a Gaussian displacement profile, which, strictly speaking, applies only to the case of unrestricted diffusion. DKI addresses this limitation by quantifying not only diffusion rate but also the degree of deviation from the Gaussian diffusion. Although the kurtosis/diffusion mismatch region showed a trend of higher perfusion level from that of the kurtosis lesion, their difference, unlike pH, was not statistically significant. This is because CBF is dynamic and under the influence of physiological conditions such as blood pressure and blood oxygenation saturation. In addition, the brain white matter and gray matter have slightly different CBF baselines. In comparison, tissue pH is reasonably stable and uniform under normal physiological conditions, making it a specific biomarker to characterize the regional tissue metabolic derangement.²³ Worth mentioning is that DKI and pH MRI results corroborate each other, substantiating the postulation that kurtosis lesion suffers worsened structural and metabolic injury than the kurtosis/diffusion lesion mismatch.²² The kurtosis/diffusion lesion mismatch, despite its DWI hyperintensity, experiences mild acidosis with pH in line with that of penumbra tissue.³² Altogether, DKI is promising to augment the perfusion/diffusion mismatch for improved mapping of ischemic penumbra; the infarction core is better defined using the kurtosis lesion and the penumbra extends into DWI lesion by including the kurtosis/diffusion lesion mismatch (Figure 4). As such, the use of stroke DKI is promising to avoid overestimation of the infarction core volume, which is crucial as the recanalization window is likely to be extended with more effective endovascular devices.

Figure 4.

DKI-aided ischemic lesion classification. (a) Routine perfusion/diffusion MRI-based tissue stratification. (b) Kurtosis MRI refines diffusion lesion mismatch into two areas: (Ia) infarction core captured by kurtosis lesion and (Ib) Viable DWI lesion captured by kurtosis/diffusion lesion mismatch.

Although it is plausible that ischemic lesion may continue to evolve during the imaging, it is likely that the initial tissue damage happens fast following filament MCAO due to its nature of severe hypoperfusion while the tissue injury stabilizes hours post-MCAO. For example, Hui et al. showed that MD, MK, and CBF lesions were stable 1–2 h after MCAO.³⁴ It has also been shown that pH-weighted MRI lesion remains stable for a few hours after MCAO.²⁵ In addition, both MD and MK were derived from the same fast DKI sequence, ensuring the same ischemia insult duration for these two indices. It is necessary to point out that the fast DKI approach proposed by Hansen et al. directly calculates MD and MK without resorting to the tensor-model based fitting.³¹ It has been shown that fast DKI and the routine tensor-based approaches provide highly correlated measurements in normal and MCAO rats.³⁵ Because the fast DKI also provides higher CNR between the intact and ischemic tissue than the tensor-model based fitting approach, it is suitable for acute stroke imaging.³⁶

Our study aims to understand the diffusion/kurtosis mismatch with multi-parametric MRI, an incremental albeit important step to improve the widely used diffusion MRI. Because pH drop induces sodium and subsequently calcium overload in the acidic tissue, causing cell injury, there may be further heterogeneity even within the kurtosis lesion.^37,38 It is necessary to note that it becomes increasingly challenging to investigate finer mismatch regions, from perfusion/diffusion mismatch, diffusion/kurtosis mismatch to possibly kurtosis/pH mismatch, and future experiments are needed to determine the spatiotemporal course of multi-parametric MRI. To our best knowledge, this is the first study to investigate the metabolic state of acute stroke DKI, which provides initial evidence that DKI refines the heterogeneous DWI lesion. Although DWI lesion reversibility is not uncommon with early reperfusion, DWI lesion often proceeds to infarction without recanalization. Because the commonly used staining techniques (e.g. hematoxylin–eosin and 2,3,5-triphenyltetrazolium chloride) are not sensitive to hyperacute ischemic tissue damage, our study used multiparametric MRI to investigate the tissue heterogeneity. It is important to point out that tissue glucose metabolic status can be assessed with autoradiographic imaging.^32,39,40 A comprehensive study that combines advanced immunohistology and novel MRI can strengthen the mechanistic understanding of acute kurtosis lesion injury to fully substantiate the refined ischemic penumbra imaging. In addition, it is necessary to evaluate advanced imaging-guided stroke treatment to test the evolution and salvageability of MRI-defined penumbra.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: NSFC (81571668 to Wu), SSTP (GJHZ20160229200622417 and JCYJ20170307165550864 to Wu), and NIH (R21NS085574 and R01NS083654 to Sun).

Acknowledgements

The authors would like to thank Gary Boas for proofreading the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors’ contributions

Drs. Wang and Wu conducted data analysis, Drs. Wang and Igarashi contributed to animal stroke model, Drs. Cheung and Zhou collected MRI data, Drs. Zhang and Sun designed the study, drafted and revised the manuscript. All authors approved the draft.

References

Moseley

Kucharczyk

Mintorovitch

et al.

Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility-enhanced MR imaging in cats. Am J Neuroradiol 1990; 11: 423–429.

Warach

Chien

et al.

Fast magnetic resonance diffusion-weighted imaging of acute human stroke. Neurology 1992; 42: 1717–1723.

Schaefer

Ozsunar

et al.

Assessing tissue viability with MR diffusion and perfusion imaging. Am J Neuroradiol 2003; 24: 436–443.

Warach

. Measurement of the ischemic penumbra with MRI: It's about time. Stroke 2003; 34: 2533–2534.

Davalos

Blanco

Pedraza

et al.

The clinical-DWI mismatch: a new diagnostic approach to the brain tissue at risk of infarction. Neurology 2004; 62: 2187–2192.

Lansberg

Thijs

Bammer

et al.

The MRA-DWI mismatch identifies patients with stroke who are likely to benefit from reperfusion. Stroke 2008; 39: 2491–2496.

Thomalla

Cheng

Ebinger

et al.

DWI-FLAIR mismatch for the identification of patients with acute ischaemic stroke within 4.5 h of symptom onset (PRE-FLAIR): a multicentre observational study. Lancet Neurol 2011; 10: 978–986.

Campbell

BCV

Mitchell

Yan

et al.

A multicenter, randomized, controlled study to investigate extending the time for thrombolysis in emergency neurological deficits with intra-arterial therapy (EXTEND-IA). Int J Stroke 2013; 9: 126–132.

Lansberg

Cereda

Mlynash

et al.

Response to endovascular reperfusion is not time-dependent in patients with salvageable tissue. Neurology 2015; 85: 708–714.

10.

Leslie-Mazwi

Hirsch

Falcone

et al.

Endovascular stroke treatment outcomes after patient selection based on magnetic resonance imaging and clinical criteria. JAMA Neurol 2016; 73: 43–49.

11.

Merino

Latour

Todd

et al.

Lesion volume change after treatment with tissue plasminogen activator can discriminate clinical responders from nonresponders. Stroke 2007; 38: 2919–2923.

12.

Yoo

Hakimelahi

Rost

et al.

Diffusion weighted imaging reversibility in the brainstem following successful recanalization of acute basilar artery occlusion. J NeuroIntervent Surg 2010; 2: 195–197.

13.

Nicoli

Lefur

Denis

et al.

Metabolic counterpart of decreased apparent diffusion coefficient during hyperacute ischemic stroke: a brain proton magnetic resonance spectroscopic imaging study. Stroke 2003; 34: e82–e87.

14.

Guadagno

Warburton

Jones

et al.

How affected is oxygen metabolism in DWI lesions? A combined acute stroke PET-MR study. Neurology 2006; 67: 824–829.

15.

Kidwell

Alger

Saver

. Evolving paradigms in neuroimaging of the ischemic penumbra. Stroke 2004; 35: 2662–2665.

16.

Labeyrie

M-A

Turc

Hess

et al.

Diffusion lesion reversal after thrombolysis: A MR correlate of early neurological improvement. Stroke 2012; 43: 2986–2991.

17.

Fiehler

Foth

Kucinski

et al.

Severe ADC decreases do not predict irreversible tissue damage in humans. Stroke 2002; 33: 79–86.

18.

Cheung

Hui

Chan

et al.

Does diffusion kurtosis imaging lead to better neural tissue characterization? A rodent brain maturation study. Neuroimage 2009; 45: 386–392.

19.

Jensen

Helpern

. MRI quantification of non-Gaussian water diffusion by kurtosis analysis. NMR Biomed 2010; 23: 698–710.

20.

Hui

Fieremans

Jensen

et al.

Stroke assessment with diffusional kurtosis imaging. Stroke 2012; 43: 2968–2973.

21.

Lavdas

Behan

Papadaki

et al.

A phantom for diffusion-weighted MRI (DW-MRI). J Magn Reson Imaging 2013; 38: 173–179.

22.

Cheung

Wang

et al.

Stratification of heterogeneous diffusion MRI ischemic lesion with kurtosis imaging – Evaluation of mean diffusion and kurtosis MRI mismatch in an animal model of transient focal ischemia. Stroke 2012; 43: 2252–2254.

23.

Hossmann

. Viability thresholds and the penumbra of focal ischemia. Ann Neurol 1994; 36: 557–565.

24.

Zhou

Payen

Wilson

et al.

Using the amide proton signals of intracellular proteins and peptides to detect pH effects in MRI. Nat Med 2003; 9: 1085–1090.

25.

Sun

Zhou

Sun

et al.

Detection of the ischemic penumbra using pH-weighted MRI. J Cereb Blood Flow Metab 2007; 27: 1129–1136.

26.

Jin

Wang

Zong

et al.

MR imaging of the amide-proton transfer effect and the pH-insensitive nuclear overhauser effect at 9.4 T. Magn Reson Med 2013; 69: 760–770.

27.

Jokivarsi

Gröhn

et al.

Proton transfer ratio, lactate, and intracellular pH in acute cerebral ischemia. Magn Reson Med 2007; 57: 647–653.

28.

Sun

Cheung

Wang

et al.

Association between pH-weighted endogenous amide proton chemical exchange saturation transfer MRI and tissue lactic acidosis during acute ischemic stroke. J Cereb Blood Flow Metab 2011; 31: 1743–1750.

29.

Sun

Wang

Cheung

. Imaging acute ischemic tissue acidosis with pH-sensitive endogenous amide proton transfer (APT) MRI – Correction of tissue relaxation and concomitant RF irradiation effects toward mapping quantitative cerebral tissue pH. Neuroimage 2012; 60: 1–6.

30.

Zhou

Guo

Igarashi

et al.

Fast diffusion kurtosis imaging (DKI) with Inherent COrrelation-based Normalization (ICON) enhances automatic segmentation of heterogeneous diffusion MRI lesion in acute stroke. NMR Biomed 2016; 29: 1670–1677.

31.

Hansen

Lund

Sangill

et al.

Experimentally and computationally fast method for estimation of a mean kurtosis. Magn Reson Med 2013; 69: 1754–1760.

32.

Peek

Lockwood

Izumiyama

et al.

Glucose metabolism and acidosis in the metabolic penumbra of rat brain. Metab Brain Dis 1989; 4: 261–272.

33.

Guo

Zhou

Chan

S-T

et al.

pH-sensitive MRI demarcates graded tissue acidification during acute stroke ― pH specificity enhancement with magnetization transfer and relaxation-normalized amide proton transfer (APT) MRI. Neuroimage 2016; 141: 242–249.

34.

Hui

Huang

et al.

Spatiotemporal dynamics of diffusional kurtosis, mean diffusivity and perfusion changes in experimental stroke. Brain Res 2012; 1451: 100–109.

35.

Sun

Wang

Mandeville

et al.

Validation of fast diffusion kurtosis MRI for imaging acute ischemia in a rodent model of stroke. NMR Biomed 2014; 27: 1413–1418.

36.

Kim

Chan

S-T

et al.

Comparison of image sensitivity between conventional tensor-based and fast diffusion kurtosis imaging protocols in a rodent model of acute ischemic stroke. NMR Biomed 2016; 29: 625–630.

37.

Siesjö

Katsura

Kristián

et al.

Molecular mechanisms of acidosis-mediated damage. Acta Neurochir Suppl 1996; 66: 8–14.

38.

Kalogeris

Baines

Krenz

et al.

Chapter six - Cell biology of ischemia/reperfusion injury. In: Kwang

(ed). International review of cell and molecular biology 2012; Vol. 298, Cambridge, MA: Academic Press, pp. 229–317.

39.

Sokoloff

Reivich

Kennedy

et al.

The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 1977; 28: 897–916.

40.

Sokoloff

. Mapping of local cerebral functional activity by measurement of local cerebral glucose utilization with [14C]deoxyglucose. Brain 1979; 102: 653–668.

pH imaging reveals worsened tissue acidification in diffusion kurtosis lesion than the kurtosis/diffusion lesion mismatch in an animal model of acute stroke

Abstract

Keywords

Introduction

Material and methods

Rodents

MRI

Data analysis

Results

Discussion

Footnotes

Funding

Acknowledgements

Declaration of conflicting interests

Authors’ contributions

References