MRI investigation of vascular remodeling for heterogeneous edema lesions in subacute ischemic stroke rat models: Correspondence between cerebral vessel structure and function

Abstract

The spatial heterogeneity in the temporal occurrence of pseudo-normalization of MR apparent diffusion coefficient values for ischemic lesions may be related to morphological and functional vascular remodeling. As the area of accelerated pseudo-normalization tends to expand faster and more extensively into the chronic stage, detailed vascular characterization of such areas is necessary. During the subacute stage of transient middle cerebral artery occlusion rat models, the morphological size of the macrovasculature, microvascular vessel size index (VSI), and microvessel density (MVD) were quantified along with functional perfusion measurements of the relative cerebral blood flow (rCBF) and mean transit time (rMTT) of the corresponding areas (33 cases for each parameter). When compared with typical pseudo-normalization lesions, early pseudo-normalization lesions exhibited larger VSI and rCBF (p < 0.001) at reperfusion days 4 and 7, along with reduced MVD and elongated rMTT (p < 0.001) at reperfusion days 1, 4, and 7. The group median VSI and rCBF exhibited a strong positive correlation (r = 0.92), and the corresponding MVD and rMTT showed a negative correlation (r = −0.48). Light sheet fluorescence microscopy images were used to quantitatively validate the corresponding MRI-derived microvascular size, density, and cerebral blood volume.

Keywords

Ischemic stroke microvessel density perfusion magnetic resonance imaging vascular remodeling vessel size index

Introduction

Following an ischemic stroke, multiple neurologic complications, including cerebral edema, inflammation, and vascular alteration, develop dynamically during the subacute stage (1 to 7 days). As the status of the subacute stage affects the progression of the chronic stage, a detailed investigation of the subacute stage of ischemic stroke is crucial. For example, pseudo-normalization of the magnetic resonance (MR) apparent diffusion coefficient (ADC) values is typically observed around one week after ischemic stroke.^1–5 This is most likely due to a combination of cytotoxic edema (lowering ADC, swelling of cells) and the development of vasogenic edema (increasing ADC, swelling of the extracellular space).^6–10 On the other hand, the temporal occurrence of such pseudo-normalization of the ADC may be spatially heterogeneous in ischemic brains.¹¹,¹² If vasogenic edema occurs faster in certain ischemic areas, an accelerated pseudo-normalization of the ADC develops, and these lesions more quickly and widely expand into the chronic stage. As a result, detailed spatial characterization and diagnosis of early and typical ADC pseudo-normalization lesions and subsequent treatment decisions may be important.⁷,^13–15 However, since these early pseudo-normalization lesions have a higher ADC (closer to ADC of normal tissue) than typical pseudo-normalization lesions in the early subacute stage (1 to 4 days), distinguishing these lesions by ADC alone can be problematic, especially in the subacute stage.

Ischemic stroke is induced by cerebrovascular dysfunction and results in vascular remodeling. The visualization and quantification of the macro- and microvascular status may help characterize and differentiate early and typical pseudo-normalization lesions. Previous MR imaging (MRI) investigations of vascular remodeling have revealed a correlation between adequate vascular remodeling, recovery, and clinical outcomes after ischemic strokes.^16–19 However, differential MRI characterizations of both vascular morphology and function between early and typical pseudo-normalization lesions are rare in the subacute stage. Consequently, simultaneous longitudinal visualization and quantification of morphological (size/density) and functional (perfusion) vascular status in both lesions can provide further insights and prognostic information in addition to general measurements of lesion tissue status via the ADC or T₂. For example, the segmentation of ischemic lesions may elucidate the origins of hyperperfusion and delayed transit time observed in the subacute stage by cross-monitoring vascular morphology and function.

As an investigation of vascular alterations after ischemic stroke, MR angiography (MRA) has shown morphological macrovascular remodeling at the brain surface region after ischemic stroke.^20–22 Microvascular remodeling in the inner brain region can be visualized by morphological mapping of the vessel size index (VSI) and microvessel density (MVD and Q), overcoming the spatial resolution limitation of MRA.^23–28 Dynamic susceptibility contrast MRI (DSC-MRI) can be used to estimate functional microvascular status, such as the cerebral blood volume (CBV), cerebral blood flow (CBF), and mean transit time (MTT) in lesions.^29–31

In this study, longitudinal morphological and functional vascular remodeling were investigated differentially between early and typical pseudo-normalization lesions to assess contrasting vascular conditions in a transient middle cerebral artery occlusion (tMCAO) rat model using dual contrast superparamagnetic iron oxide nanoparticles (SPION).²⁸ Early and typical pseudo-normalization lesions were selected and segmented using longitudinally acquired ADC maps. Corresponding VSI and MVD maps were generated to evaluate the morphological vascular status. For evaluation of functional vascular status, relative CBV (rCBV), CBF (rCBF), and MTT (rMTT) maps from DSC-MRI acquisitions were simultaneously analyzed. Ultra-short echo time MRAs (UTE-MRAs) were acquired to compare morphological macrovascular remodeling occurring at the brain surface region. For the direct validation of VSI, MVD, and rCBV values between two lesions, light sheet fluorescence microscopy (LSFM) images of tMCAO rat models were acquired and quantitatively compared at reperfusion day 7.

Material and methods

Animal preparation

Animal experiments were performed in accordance with the Animal Protection Act of Korea under a protocol approved by the Institutional Animal Care and Use Committee of the Ulsan National Institute of Science and Technology. All experiments were reported in compliance with the ARRIVE guidelines (Animal Research: Reporting In Vivo Experiments).³² Male Wistar rats (bodyweight 280-320 g, n = 13) were subjected to focal brain ischemia by transient occlusion of the middle cerebral artery (MCA). Initially, rats were anesthetized by inhalation of 3% isoflurane in a mixture of 70% nitrous oxide and 30% oxygen. Anesthesia was maintained by changing the range of isoflurane to 1-1.5%. Body temperature was kept constant (37 $\pm$ 1 $° C$ ) through a feedback-controlled heating pad. As previously described,³³ an intraluminal monofilament (0.37 mm diameter, Doccol Corporation, Redlands, CA, USA) was inserted into the external carotid artery and positioned to the internal carotid artery to block the MCA. The monofilament was withdrawn after 1 h of occlusion, and rats were housed in a 12 h light-12 h dark cycle with access to food and water ad libitum.

MRI acquisition

MRI acquisitions were conducted using a 7-T MR scanner (Bruker, Ettlingen, Germany). The tMCAO rat models were anesthetized using the same method of focal brain ischemia surgery. A single contrast agent, SPION, was synthesized in-house according to previously described methods.³⁴ The core size distribution of the iron oxide measured by transmission electron microscopy was 5–10 nm, and the mean hydrodynamic diameter of the iron oxide nanoparticles measured by differential light scattering was 20 $\pm$ 7 nm. The r₁ and r₂ of SPION were 2.36 mM⁻¹s⁻¹ and 32.94 mM⁻¹s⁻¹ at 7-T, respectively. SPION was administered as an intravenous bolus with a dose of 75 $μ$ mol Fe/Kg for DSC-MRI acquisition, and an additional SPION dose of 285 $μ$ mol Fe/Kg (total 360 $μ$ mol Fe/Kg) was administered for $Δ$ R₂, $Δ$ R₂*, and UTE-MRA acquisitions. Longitudinal MRI acquisitions of tMCAO rat models were performed at reperfusion days 1, 4, and 7. Morphological and functional microvascular remodeling (VSI and MVD and DSC-MRI, respectively) were acquired for all (n = 11) tMCAO rat models, and macrovascular remodeling (UTE-MRA) was acquired for two representative tMCAO rat models showing early pseudo-normalization and typical pseudo-normalization in the cortex lesions.

The ADC map was acquired using a diffusion-weighted echo planar imaging (EPI) pulse sequence with the following parameters: TR/TE = 3500/26.6 ms, number of averages (NA) = 8, number of segments = 3, b-values = 100, 200, 400, 600, 800, and 1000 s/mm², flip angle (FA) = 90 $°$ , matrix size = 100 $\times$ 100, field of view (FOV) = 30 $\times$ 30 mm², resolution = 300 $\times$ 300 $μ$ m², number of slices = 8, and slice thickness = 1 mm. Three directional (x, y, and z) ADC maps were acquired and averaged.

The DSC-MRI perfusion maps were acquired using a gradient-echo EPI pulse sequence with the following parameters: TR/TE = 420/17 ms, NA = 1, number of segments = 1, number of repetitions = 400, FA = 30 $°$ , matrix size = 96 $\times$ 96, FOV = 30 $\times$ 30 mm², resolution = 313 $\times$ 313 $μ$ m², number of slices = 8, and slice thickness = 1 mm.

The $Δ$ R₂ map was acquired using a multi-slice multi-echo pulse sequence with the following parameters: TR = 6000 ms, TE = 8-160 ms, echo spacing = 8 ms, NA =1, FA = 90 $°$ , matrix size = 256 $\times$ 256, FOV = 30 $\times$ 30 mm², resolution = 117 $\times$ 117 $μ$ m², number of slices = 8, and slice thickness = 1 mm.

The $Δ$ R₂^* map was acquired using a multi-echo gradient-echo pulse sequence with the following parameters: TR = 6000 ms, TE = 3-59 ms, echo spacing = 4 ms, NA = 1, FA = 90 $°$ , matrix size = 256 $\times$ 256, FOV = 30 $\times$ 30 mm², resolution = 117 $\times$ 117 $μ$ m², number of slices = 8, and slice thickness = 1 mm.

The UTE-MRA was acquired using the UTE pulse sequence with the following parameters: TR/TE = 22/0.012 ms, NA = 1, FA = 40 $°$ , under-sampling factor = 1.13, matrix size = 512 $\times$ 512 $\times$ 512, FOV = 30 $\times$ 30 $\times$ 30 mm³, and resolution = 59 $\times$ 59 $\times$ 59 $μ$ m³.

Fluorescence image acquisition

To label cerebral blood vessels, fluorescein isothiocyanate (FITC)-dextran solution was heart perfused in two representative tMCAO rat models, which showed early pseudo-normalization and typical pseudo-normalization in the cortex lesions at reperfusion day 7 according to the previously described methods.³⁵ One hundred milliliters of saline solution was administered into the left ventricle to extract cerebrovascular blood through the right atrium. Subsequently, FITC-dextran was mixed with 5% gelatin solution at a concentration of 0.075% and administered into the left ventricle. After 30 min of refrigeration, the rat brain was dissected and fixed in 4% paraformaldehyde (PFA) for two days. Then, 2 mm of the brain, including the edema lesion, was dissected. Tissue clearing was then performed to acquire 3-dimensional structural images using a Binaree tissue clearing™ kit (Cat. HRTC-001; Binaree, Korea). In brief, the tissue was immersed in the Binaree tissue clearing solution, which was changed every two to three days until the sample was visibly clear. The brains were transferred to Binaree Mounting and Storage™ Solution (Cat. SHMS-060, Binaree) for refractive index matching. Subsequently, using LSFM (Zeiss Lightsheet Z.1), 3-dimensional fluorescence images of tissue-cleared tMCAO rat brains were acquired with a resolution of 1.2 $\times$ 1.2 $\times$ 5.2 $μ$ m³. In addition, 4% PFA fixed brains were embedded in the OCT compound and dissected using a cryostat (Leica) at a thickness of 100 $μ$ m. Fluorescence images were acquired at a resolution of 0.8 $\times$ 0.8 $μ$ m² using a fluorescence stereo zoom microscope (Zeiss Axio Zoom. V16) with a 38 Zeiss filter (excitation wavelength of 470 $\pm$ 40 nm and an emission wavelength of 525 $\pm$ 50 nm).

Data processing and analysis

MATLAB (MathWorks, Natick, MA) and ImageJ (US National Institutes of Health, Bethesda, MD) software were used for the data processing and analysis. Voxel-wise ADC values were estimated by fitting the equation S = S₀ $\times e^{- ADC \times b}$ with a nonlinear least-squares fitting method using MATLAB. b refers to the b-value for the corresponding MRI acquisition. Acquired ADC maps were up-sampled to a size of 256 $\times$ 256 for VSI and MVD calculations. rCBV, rMTT, and rCBF maps were acquired using DSC-MRI. Initially, time-series signal intensity alteration due to SPION passage of each voxel was measured to calculate $Δ$ R₂^*(t) = $- \frac{1}{TE} l n (\frac{S (t)}{S_{pre}})$ , where S_pre is an averaged signal before the administration of SPION. Subsequently, the area under the curve of $Δ$ R₂^*(t) was measured after fitting $Δ$ R₂^*(t) with the gamma-variate function to calculate the indices of rCBV, MTT, and rCBF using rCBV_index = $\int_{0}^{t} Δ$ R₂^*( $τ$ )d $τ$ , MTT_index = $\frac{\int_{0}^{t} τ Δ {R_{2}}^{*} (τ) d τ}{\int_{0}^{t} Δ {R_{2}}^{*} (τ) d τ}$ , rCBF_index = $\frac{{rCBV}_{index}}{{MTT}_{index}}$ .³⁶ Finally, each rCBV, rMTT, and rCBF map was generated by dividing the index by the median value of the contralateral corpus callosum region indices for each tMCAO rat model.

The $Δ$ R₂ and $Δ$ R₂^* values were calculated by subtracting the transverse relaxation rate (R₂ and R₂^*) values acquired before the administration of SPION from the respective values acquired after the administration of SPION. The voxel-wise R₂ and R₂^* values were estimated by fitting the equation S = S₀ $\times e^{- TE \times [R_{2} or {R_{2}}^{*}]}$ . $Δ$ R₂ and $Δ$ R₂^* maps were used to determine the VSI and MVD maps using the following equations³⁷,³⁸

VSI (μm) =0 .424 {(ADC / ({γΔχB}_{0}))}^{1 / 2} {({ΔR}^{*}_{2} / {ΔR}_{2})}^{3 / 2}

(1)

MVD ({mm}^{- 2}) \approx Q^{3} / (4 .725ADC)

(2)

where

γ

is the proton gyromagnetic ratio,

Δ χ

is the magnetic susceptibility difference between the vessel and tissue, and B₀ is the magnetic field strength.

Δ χ

was estimated by

Δ χ

\frac{3 Δ {R_{2}}^{*}}{4 π BVf γ B 0}

, where BVf is the blood volume fraction. In addition, the microvessel density index Q was estimated by Q (s^−1/3) =

Δ

R₂/(

Δ

R₂^*)^2/3.

Acquired UTE-MRAs were denoised and visualized using 3 D Slicer software (www.slicer.org) to investigate macrovascular remodeling in the tMCAO rat models.²⁸ The caudal rhinal veins (cRHVs), populated in the contralateral and ipsilateral regions for each representative tMCAO rat model, were selected and segmented for macrovascular remodeling. Vessel diameters of the segmented cRHVs were calculated using BoneJ software integrated with ImageJ.³⁹ In addition, the SD of each cRHV was derived.

Due to the DSC-MRI-derived rCBV maps’ low spatial resolution, rCBV maps were replaced with high spatial resolution $Δ$ R₂^* maps. The values of the ADC, T₂, VSI, MVD, CBV ( $Δ$ R₂^*), rCBF, and rMTT in early pseudo-normalization lesions, typical pseudo-normalization lesions, and contralateral regions were calculated and quantitatively evaluated. Subsequently, the relationship of vascular remodeling according to the status of cerebral edema was compared. A Mann-Whitney U test was performed to compare significant differences between early and typical pseudo-normalization lesions. To validate the difference in MRI-derived microvascular remodeling between early and typical pseudo-normalization lesions at reperfusion day 7, the vessel radius, vessel density, and blood volume fraction from the LSFM images were evaluated. The LSFM images were resized to an isotropic resolution of 1 $\times$ 1 $\times$ 1 $μ$ m³. Three different regions of interest (ROIs) were selected for each early pseudo-normalization, typical pseudo-normalization lesions, and contralateral regions (total 12 ROIs). Then, the vessel radii were measured using BoneJ software with the same process as the vessel diameter calculation of UTE-MRA. The vessels in the segmented ROIs were counted using ImageJ software to calculate the vessel density.

Results

Early pseudo-normalization and typical pseudo-normalization lesions

Even though the pathogenesis of cerebral edema after ischemic stroke is complex, the underlying ADC dependence of early pseudo-normalization and typical pseudo-normalization lesions with associated vascular remodeling are conceptualized in Figure 1. To address the heterogeneity in the pseudo-normalization of ADC in ischemic lesions, two different edema progressions, marked by gold and green solid lines, are shown in Figure 1(a). The gold line shows the pseudo-normalization of ADC before reperfusion day 7 as marked by the gold arrow (early pseudo-normalization), while the green line shows the pseudo-normalization of ADC at reperfusion day 7 as marked by the green arrow (typical pseudo-normalization). The gold line shows a much higher ADC than the green line at reperfusion day 7. In Figure 1(b), the exemplar MRI-derived morphological VSI map and functional rCBF map are both visualized along with volume-rendered LSFM images, which were quantitatively analyzed and compared between the lesions of early and typical pseudo-normalization of ADC in the current study.

Figure 1.

Conceptual illustration of edema progression in typical pseudo-normalization and early pseudo-normalization lesions. (a) ADC alterations in typical pseudo-normalization (green solid line) and early pseudo-normalization (gold solid line) lesions as reperfusion continued from day 1 to 7. The black dashed line indicates the ADC value of normal tissue. Two different time points of ADC pseudo-normalization in early and typical pseudo-normalization lesions are marked by gold and green arrows, respectively. (b) Exemplar MRI-derived morphological vascular remodeling related VSI map, functional vascular remodeling related rCBF map, and volume-rendered LSFM images to characterize vascular remodeling in typical pseudo-normalization and early pseudo-normalization lesions.

Morphological and functional microvascular remodeling mapping

Overall, Figure 2 shows longitudinally (reperfusion days 1, 4, and 7) acquired edema status surrogate (T₂ and ADC) maps with morphological (VSI and MVD) and functional (CBV, rCBF, and rMTT) microvascular maps of a representative tMCAO rat model. At reperfusion day 1, the area of increased T₂ with decreased ADC indicates the edema lesion, as marked by yellow arrows is shown in Figure 2(a) and (b), respectively. Correspondingly, morphological (increased VSI and decreased MVD) and functional (reduced rCBF and elongated rMTT) microvascular variations are spatially matched with the corresponding edema lesion as shown in Figure 2(c), (d), (f), and (g), respectively. At reperfusion day 4, a slight decrease in T₂ and an increase in ADC values with respect to those values at reperfusion day 1 were observed, as shown in Figure 2(h) and (i), respectively. Along with such edema progression, spatially heterogeneous morphological and functional microvascular alterations are apparent (Figure 2(j) to (n)). Specifically, there are conspicuous lesions marked by red arrows in Figure 2(h) and (i), which exhibit significantly increased T₂ and ADC. Such areas represent so-called early pseudo-normalization lesions. Even at reperfusion day 7, early pseudo-normalization lesions (ADC > 1500 $μ$ m²/s at reperfusion day 7) were observed with persistently enlarged VSI and reduced MVD values with respect to those values at reperfusion day 4, as shown in Figure 2(p) to (r). On the other hand, typical pseudo-normalization (ADC ∼ 800 $μ$ m²/s at reperfusion day 7) tended to restore the morphological and functional microvascular parameters (except for rMTT), as shown in Figure 2(p) to (t).

Figure 2.

The representative lesion tissue status, as well as morphological and functional microvascular remodeling maps of one tMCAO rat model. (a–u) Longitudinally (at reperfusion days 1, 4, and 7) acquired lesion tissue status related T₂ and ADC maps, morphological microvascular remodeling related VSI and MVD maps, and functional microvascular remodeling related CBV, rCBF, and rMTT maps of one tMCAO rat model. The yellow arrows indicate edema lesion and red arrows indicate early pseudo-normalization lesions.

ROI selection for early and typical pseudo-normalization lesions and the contralateral region

According to the ADC maps, typical pseudo-normalization, early pseudo-normalization lesions, and the contralateral region were selected and segmented, as shown in Figure 3. To cover whole-brain cerebral edema, three or four slices of each ADC map were used. At reperfusion day 1, ADC values under 650 $μ$ m²/s were considered as edema lesions and segmented as typical pseudo-normalization lesions, as indicated by the yellow boundary line in Figure 3(a). Then, the mirror region in the contralateral hemisphere was segmented as the contralateral region, as indicated by the green boundary line in Figure 3(a). Based on the segmented typical pseudo-normalization lesion and the contralateral region at reperfusion day 1, the same ROIs were registered to the ADC maps of reperfusion days 4 and 7 and segmented (Figure 3(b) and (c)). In addition, vasogenic edema dominant and early pseudo-normalization lesions were selected and segmented. At reperfusion day 7, ADC values over 1500 $μ$ m²/s were considered as vasogenic edema dominant and segmented as early pseudo-normalization lesions, as indicated by the red boundary line in Figure 3(c). As contralateral tissue regions had ADC values on the order of 800–850 $μ$ m²/s, ADC values below 650 $μ$ m²/s at reperfusion day 1 and above 1500 $μ$ m²/s at reperfusion day 7 discriminated typical and early pseudo-normalization lesions from normal tissue, respectively. The ventricle regions show ADC values over 1500 $μ$ m²/s and if they are included in early pseudo-normalization lesions, they can induce bias to various vascular remodeling parameters. However, as accelerated vasogenic edema induced early pseudo-normalization lesions are structurally distinct from ventricle regions, the ventricle regions were not included in early pseudo-normalization lesions. Furthermore, T₂ values in early pseudo-normalization lesions were conspicuously lower compared to those of ventricle regions, T₂ maps further confirmed the difference between early pseudo-normalization lesions and ventricular regions. Based on the segmented early pseudo-normalization lesion at reperfusion day 7, the same ROI was registered to the ADC maps of reperfusion days 1 and 4. From the mask of typical pseudo-normalization ROIs, the lesions of early pseudo-normalization ROIs were excluded for all reperfusion days. In addition, at each reperfusion day, the mirror region of the early pseudo-normalization lesion in the contralateral region was included as the contralateral region. All segmented ROIs were registered in DSC-MRI-derived perfusion maps as well.

Figure 3.

Respective ROIs for a typical pseudo-normalization lesion, early pseudo-normalization lesion, and contralateral region. (a–c) Selected ROIs for the typical pseudo-normalization lesion (yellow boundary lines, ADC < 650 $μ$ m²/s at reperfusion day 1), early pseudo-normalization lesion (red boundary lines, ADC > 1500 $μ$ m²/s at reperfusion day 7), and contralateral region (green boundary lines) at reperfusion day 1, 4, and 7. The contralateral region was selected as mirror regions of both typical and early pseudo-normalization lesions.

Comparisons of the morphological and functional microvascular remodeling

Next, a total of 33 cases of each of the ADC, T₂, and morphological and functional microvascular parameters were longitudinally (reperfusion day 1, 4, and 7) mapped for the 11 tMCAO rat models. Their median values were compared among the three ROIs, that is, typical and early pseudo-normalization lesions and the contralateral region. Figure 4 shows violin plots (distributions, median values, and interquartile ranges) of the edema status (ADC and T₂), as well as the morphological (VSI and MVD) and functional (CBV, rCBF, and rMTT) microvascular alterations of typical (blue) and early (red) pseudo-normalization lesions according to reperfusion days 1, 4, and 7. The median values are connected with dashed lines to evaluate the trend of vascular remodeling. In addition, the median values of the contralateral regions are connected with solid lines (orange) as reference values. In ADC, early pseudo-normalization lesions exhibited a median ADC value comparable to that of the contralateral region around reperfusion day 2, while typical pseudo-normalization lesions displayed it around reperfusion day 7 (Figure 4(a)). Elevated median T₂ values in typical and early pseudo-normalization lesions decreased over time (Figure 4(b)), except that early pseudo-normalization lesions showed increasing median T₂ values as reperfusion continued from day 4 to 7. For morphological microvascular alterations, early pseudo-normalization lesions showed higher median VSI values than typical pseudo-normalization lesions at reperfusion days 4 and 7, while similar values were observed at reperfusion day 1 (Figure 4(c)). In addition, early pseudo-normalization lesion showed increased macrovascular diameter in UTE-MRAs compared to typical pseudo-normalization lesion as shown in Supplementary Figure 1. In Supplementary Figure 1(q) and (r), the estimated cRHV diameters are shown as bar graphs. Both of the tMCAO rat models showed no significant alteration of cRHVs in the contralateral regions (orange) as reperfusion continued until day 7. In typical pseudo-normalization lesion, enlarged cRHV at reperfusion day 4 decreased at reperfusion day 7, indicating normalization of the macrovasculature (blue). However, in early pseudo-normalization lesion, enlarged cRHV at reperfusion day 4 did not significantly decrease at reperfusion day 7 compared to typical pseudo-normalization lesions (red). Early pseudo-normalization lesions showed lower median MVD values than typical pseudo-normalization lesions at all reperfusion days (Figure 4(d)). However, typical pseudo-normalization lesions showed increased median MVD values as reperfusion continued from day 4 to 7, while early pseudo-normalization lesions showed persistently lowered median MVD values. For functional microvascular alterations, early pseudo-normalization lesions showed higher median CBV and rCBF values than typical pseudo-normalization lesions at reperfusion days 4 and 7 (Figure 4(e) and (f), respectively). Early pseudo-normalization lesions showed higher median rMTT values than typical pseudo-normalization lesions at all reperfusion days (Figure 4(g)). As reperfusion continued from day 4 to 7, the median values of rMTT in early pseudo-normalization lesions became further elongated. For detailed distribution comparisons between typical pseudo-normalization lesions and contralateral regions, violin plots are shown in Supplementary Figure 2. After Anderson-Darling tests for normal distribution decision, all Mann-Whitney U tests showed significant differences between typical pseudo-normalization and early pseudo-normalization lesions (p < 0.01) except for T₂ and VSI at reperfusion day 1 (p = 0.371 and p = 0.947, respectively).

Figure 4.

Violin plots of the lesion tissue status as well as morphological and functional microvascular remodeling. (a–g) Violin plots (distributions, median values, and interquartile ranges) of the lesion tissue status related ADC and T₂, morphological microvascular remodeling related VSI and MVD, and functional microvascular remodeling related CBV, rCBF, and rMTT in typical pseudo-normalization (blue) and early pseudo-normalization (red) lesions at reperfusion days 1, 4 and 7. The solid orange lines indicate the connected median values of the contralateral regions. The median values of each typical pseudo-normalization and early pseudo-normalization lesion were connected with dashed lines as reperfusion continued from days 1 to 7. The X marks indicate that no significant difference between typical and early pseudo-normalization lesions was found in the statistical analysis (T₂ and VSI at reperfusion day 1, p = 0.371 and p = 0.947 from the Mann-Whitney U tests, respectively).

Correlations between the VSI versus rCBF and the MVD versus rMTT

Scatter plots of the VSI versus rCBF for hyperperfused and the MVD versus rMTT for transit-delay regions are shown in Figure 5 to investigate the relationships between morphological and functional microvascular parameters. The results especially link hyperperfusion versus vessel size and delayed transit time versus vessel density. Individual median VSI, rCBF, MVD, and rMTT values were acquired from three ROIs (typical pseudo-normalization, early pseudo-normalization lesions, and the contralateral region) of each tMCAO rat model for each reperfusion day and are shown in a scatterplot in Figure 5(a) and (b) (cases with hyperperfusion (a) and delayed transit time (b) were considered). Subsequently, the group median VSI, rCBF, MVD, and rMTT values of all tMCAO rat models were scatter-plotted, as shown in Figure 5(c) and (d). The group median values of typical pseudo-normalization (triangle) and early pseudo-normalization (star) lesions were separately included in scatterplots. After Anderson-Darling tests for normal distribution decision, an individual comparison of the VSI and rCBF showed a positive correlation (r = 0.454, p < 0.001 from paired student’s t-test) (Figure 5(a)), while the group comparison showed a strong positive correlation (r = 0.92, p < 0.001 from paired student’s t-test) (Figure 5(c)). Increased vessel size appears to be related to observed hyperperfusion in subacute ischemic lesions. On the contrary, an individual comparison of the MVD and rMTT showed a negative correlation (r = −0.415, p < 0.001 from Wilcoxon signed-rank test) (Figure 5(b)), and the group comparison also showed a negative correlation (r = −0.477, p = 0.012 from paired student’s t-test) (Figure 5(d)). Individual comparison of the MVD and rMTT (blue circles) showed a much higher correlation at reperfusion day 7 than that at reperfusion days 1 (red circles) and 4 (green circles) with r = -0.604 and p < 0.001 from Wilcoxon signed-rank test (inset figure of Figure 5(b)). This was presumably due to the distinct differences in the MVD and rMTT between early and typical pseudo-normalization lesions at reperfusion day 7. In addition, from the group median comparison, increasing rMTT values were pair-wisely observed for early pseudo-normalization lesions with decreasing MVD values compared to those of typical pseudo-normalization lesions for each reperfusion day.

Figure 5.

Correlations between the VSI versus rCBF and the MVD versus rMTT. (a, b) Scatter plots of the median rCBF and rMTT values over the averaged median rCBF and rMTT values in the contralateral regions of each tMCAO rat model with the corresponding median VSI and MVD values according to reperfusion day 1 (red circles), 4 (green circles), and 7 (blue circles). (inset figure of (b)) Scatter plot of MVD and rMTT at reperfusion day 7. Correlation coefficient r = 0.454, r = −0.415, and r = −0.604 and p-values < 0.001 from the paired student’s t-test and Wilcoxon signed-rank tests, respectively. (c, d) Scatter plots of the median VSI versus rCBF and the MVD versus rMTT values of typical pseudo-normalization (triangle) and early pseudo-normalization lesions (star) from all tMCAO rat models according to reperfusion day 1 (red), 4 (green), and 7 (blue). Correlation coefficient r = 0.92 and r = −0.477 and p < 0.001 and p = 0.012 from the paired student’s t-tests, respectively. The black solid lines indicate the fitted lines. The orange dashed lines indicate the average median rCBF and rMTT values of the contralateral regions.

Comparisons of the MRI- and LSFM-derived microvascular parameters

To validate the microvascular parameters of the VSI, MVD, and CBV ( $Δ$ R₂^*) derived from MRI, LSFM images of two representative tMCAO rat models exhibiting typical pseudo-normalization and early pseudo-normalization in cortex lesions were acquired and quantitatively compared. The contralateral regions and ipsilateral lesions were separately compared in Figures 6 and 7, respectively. For each tMCAO rat model, six ROIs were selected and analyzed. Supplementary Figure 3(a) shows six representative ROIs (three red ROIs in the ipsilateral lesion and three blue ROIs in the contralateral region) in the maximum intensity projection (MIP) image of 64 $μ$ m thickness LSFM images. In addition, a volume-rendered LSFM image via Imaris software (Bitplane, Switzerland) at a thickness of 2 mm is shown for reference in Supplementary Figure 3(b) and addresses the 3-dimensional nature of LSFM images.

Figure 6.

Quantitative comparisons of the MRI- and LSFM-derived vessel size and vessel density in the contralateral cortex regions. (a, b) Exemplar VSI and MVD maps at reperfusion day 7. The yellow arrows and overlapped yellow regions in each map indicate the segmented ROI in the contralateral cortex region for comparison. (c, d) Plots of the MRI-derived median VSI and MVD values in the contralateral cortex region for each tMCAO rat model. (e, f) MIP images of 64 $μ$ m thickness LSFM images of two representative tMCAO rat models and two ROIs (blue boxes) with zoomed-in volume-rendered images. (g, h) Plots of the LSFM-derived mean vessel radius and vessel density at each ROI. The red solid lines indicate the average values of the vessel sizes and densities derived from MRI and LSFM. (i–k) Comparison of the average mean vessel radii, vessel densities, and BVfs of six ROIs (three ROIs for each tMCAO rat model) according to the thickness of the LSFM image. The red arrows indicate the thickness of the LSFM image for the mean vessel radius and vessel density calculation and comparison (g, h, respectively).

Figure 7.

Quantitative comparisons of the MRI- and LSFM-derived vessel size, vessel density, and blood volume in the ipsilateral cortex lesions. (a, b) MIP images of 64 $μ$ m thickness LSFM images of early pseudo-normalization and typical pseudo-normalization in different tMCAO rat models. (c, d) Zoomed-in volume-rendered images of ROIs (red boxes) in (a) and (b). (e, f) T₂-weighted images at reperfusion day 1. (g, h) ADC maps at reperfusion day 7. (i–k) Comparison of the LSFM-derived averaged mean vessel radii, vessel densities, and BVfs of three ROIs according to the thickness of LSFM image in early pseudo-normalization (green dashed lines) and typical pseudo-normalization lesions (yellow dashed lines). (l–n) The ratios of vessel size, vessel density, and blood volume according to the LSFM image thickness between typical pseudo-normalization (TP) and early pseudo-normalization (EP) lesions. The red arrows indicate the ratios of vessel size (VSI), vessel density (MVD), and blood volume (CBV) derived from MRI.

MRI-derived morphological microvascular parameters of the VSI and MVD values in the contralateral cortex regions were acquired through six slices at reperfusion day 7 for each tMCAO rat model, and an exemplar ROI is shown as a yellow region overlapped on the VSI and MVD maps, as shown in Figure 6(a) and (b), respectively. The median values of the VSI and MVD of each tMCAO rat model are shown as green dashed lines in Figure 6(c) and (d), respectively, for 11 cases. The red lines show the averaged VSI and MVD values of 11 tMCAO rat models, which are 4.2 $μ$ m and 314.2 mm⁻², respectively. MIP images of 64 $μ$ m thickness LSFM images are shown in Figure 6(e) and (f). In addition, two ROIs (blue boxes) in the contralateral cortex regions with zoomed-in volume-rendered images of each ROI are shown. The mean vessel radius and vessel density were calculated at six ROIs in the contralateral cortex regions (three ROIs for each tMCAO rat model) as blue dashed lines in Figure 6(g) and (h), respectively. The red lines show the average values of the mean vessel radius and vessel density, which are 4.26 $μ m$ and 286.3 mm⁻², respectively. The mean vessel radius and vessel density in Figure 6(g) and (h) were calculated from the thickness of 256 $μ$ m and 4 $μ$ m, respectively, as marked with red arrows in Figure 6(i) and (j), respectively. In the contralateral cortex region, the MRI-derived average VSI and MVD values and LSFM-derived average values of the mean vessel radius and vessel density were quantitatively comparable. To further investigate the thickness dependence of the LSFM-derived parameters, the mean vessel radius, vessel density, and BVf were calculated according to the LSFM image thicknesses of 4, 8, 16, 32, 64, 128, and 256 $μ$ m, as shown in Figure 6(i) to (k), respectively. The mean vessel radius, vessel density, and BVf of the six ROIs were averaged at each thickness. As the thickness increased, the vessel radius (∼ 4 $μ$ m) and BVf (∼ 2.5%) were not significantly altered, while the vessel density increased, which is also directly visible in the volume-rendered images of one ROI in the contralateral region according to various thicknesses as shown in Supplementary Figure 4.

Next, the ipsilateral cortex lesions of two representative cases were compared, as shown in Figure 7. MIP images of 64 $μ$ m thickness LSFM images are shown in Figure 7(a) and (b). In addition, two ROIs (blue boxes) in the ipsilateral cortex lesions with zoomed-in volume-rendered images of each ROI are shown in Figure 7(c) and (d). Figure 7(c) visually shows a larger vessel size and lower vessel density compared to those of Figure 7(d). At reperfusion day 1, cerebral edema was observed in both T₂-weighted images (Figure 7(e) and (f)). However, at reperfusion day 7, the ADC map showed abnormally increased ADC values compared to those of the contralateral regions in Figure 7(g), which is a signature of vasogenic edema dominant and early pseudo-normalization lesions. The other tMCAO rat model showed a typical pseudo-normalization case (Figure 7(h)). For LSFM images of thicknesses of 4, 8, 16, 32, 64, 128, and 256 $μ$ m, the average mean vessel radius, vessel density, and BVf of three ROIs for each tMCAO rat model are shown in Figure 7(i) to (k), respectively. The early pseudo-normalization lesion showed a higher mean vessel radius than the typical pseudo-normalization lesion for all thicknesses (Figure 7(i)). A similar trend was observed for the BVf comparison in Figure 7(k). The early pseudo-normalization lesion showed a lower vessel density than the typical pseudo-normalization lesion for all thicknesses (Figure 7(j)). For quantitative comparison between the LSFM- and MRI-derived vascular parameters (size, density, and blood volume), the ratios between typical and early pseudo-normalization lesions were compared, as shown in Figure 7(l) to (n), respectively. The ratios of MRI-derived vascular parameters were calculated from the median values of all tMCAO rat models at reperfusion day 7. The LSFM-derived ratios were in comparable ranges compared to those that were MRI-derived as marked by red arrows.

Fluorescence images acquired using a fluorescence stereo zoom microscope (without tissue clearing) also verified morphological microvascular remodeling differences between early and typical pseudo-normalization lesions, as shown in Supplementary Figure 5. Supplementary Figure 5(g), (i), (h), and (j) show zoomed-in images of ROIs in typical pseudo-normalization lesion (blue box), early pseudo-normalization lesion (red box), and contralateral regions (orange boxes), respectively. The zoomed-in images show increased vessel size and decreased vessel density in early pseudo-normalization lesion compared to typical pseudo-normalization lesion.

Discussion

After a week of reperfusion, tMCAO rat models showed typical pseudo-normalization lesions in the ADC maps. However, there were spatiotemporally different lesions showing early pseudo-normalization of the ADC before then. Accordingly, typical pseudo-normalization and early pseudo-normalization lesions were separately investigated using the edema status (ADC and T₂), and morphological (VSI and MVD), and functional (rCBF and rMTT) vascular remodeling longitudinally. It should be noticed that focal brain ischemia was systematically induced without monitoring of sequential ADC or CBF, which may result in potential bias to acquired data because of inconsistently occluded MCA. At reperfusion day 1, the tMCAO rat model showing no edema lesion in T₂-weighted image and/or ADC map was excluded from this study. Although heterogeneous morphological and functional microvascular remodeling were observed, correlative responses between them during the progression of ischemic stroke have not been well established. In a clinical study, hyperperfusion was observed after recanalization, and there is controversy regarding its effectiveness in hyperperfusion. Some cerebral hemodynamic studies have reported that hyperperfusion is a sign of successful recanalization with better clinical outcomes,^40–43 while others have reported that hyperperfusion is correlated with hemorrhagic transformation.⁴⁰,⁴⁴,⁴⁵ It should be considered that in this study, early pseudo-normalization lesions showed higher rCBF values compared to typical pseudo-normalization lesions. As early pseudo-normalization lesions have a dominance of vasogenic edema, such lesions are more likely to accompany the damaged blood-brain barrier. Thus, hyperperfusion observed in the subacute stage requires cautionary interpretation. The relationship between vascular size and blood volume (flow) in the subacute stage is also informative. The morphological vascular alteration was observed to regulate the CBF,⁴⁶,⁴⁷ and this study highlights that hyperperfusion in the subacute stage is driven by enlarged vessel size, even with reduced vessel density and delayed transit time. Besides hyperperfusion, delayed transit time was also observed in lesions. A previous study reported that the net extraction of oxygen negatively correlates with the capillary transit time and is limited mainly by capillary density.⁴⁸ The observed delayed transit time and low microvessel density may implicate delayed metabolism due to delayed oxygen and nutrient supply to the lesion. In addition, the association of delayed transit time with respect to low microvessel density provides insight into the role of capillaries in regulating (limiting) the CBF. Although confounding correlations may exist between morphological and functional microvascular remodeling, this result emphasizes the importance of monitoring morphological and functional vascular remodeling after ischemic stroke. Further investigation of vascular remodeling and lesion tissue status is required to estimate the recovery or progression of ischemic stroke in the acute stage.

It is important to emphasize that early pseudo-normalization lesions showed faster and more extensive expansion in the chronic stage. For example, in Supplementary Figure 6, two tMCAO rat models showed different vasogenic edema dominancy in lesions at reperfusion day 7 (Supplementary Figure 6(e) and (f)). A tMCAO rat model showing extensive vasogenic edema dominant and early pseudo-normalization lesion at reperfusion day 7 displayed much faster expansion of early pseudo-normalization lesions at the chronic stage (reperfusion days 16 and 25), as shown in Supplementary Figure 6(g) to (j). In connection with the aggravation of early pseudo-normalization lesion, unnormalized morphological and functional vascular remodeling were observed compared to those of typical pseudo-normalization lesion. This emphasizes the necessity to characterize the lesions of early pseudo-normalization in the early subacute stages. This study observed that reduced MVD and elongated rMTT values were more discriminative factors of early pseudo-normalizations lesions at reperfusion day 1 than conventional ADC and T₂ values, as shown in Figure 4 at reperfusion day 1. Interquartile ranges of MVD and rMTT in early pseudo-normalization lesion showed the most conspicuous difference compared to those of typical pseudo-normalization lesion except for the ADC whose interquartile range of early pseudo-normalization lesion is close to the median value of the contralateral region. Therefore, it can be difficult to distinguish early pseudo-normalization lesion from normal tissue at reperfusion day 1 with ADC values alone. However, in the early pseudo-normalization lesion, CBV and rCBF showed more conspicuous alterations compared to rMTT in Figure 4, which required further investigation regarding their association with edema progression. Also, further investigation of CBV and rCBF as discriminative factors for differentiating early pseudo-normalization lesions from typical pseudo-normalization lesions may be useful. As the two lesions showed different tendencies for vascular remodeling and edema progressions, assessing morphological and functional vascular remodeling in both edema lesions up to the chronic stage could provide more accurate diagnostic and therapeutic information of ischemic stroke.

To validate MRI-derived microvascular remodeling related to the VSI, MVD, and CBV, 3-dimensional LSFM images were acquired and analyzed. In the contralateral cortex regions, the MRI- and LSFM-derived vessel size (4.2 $μ$ m and 4.26 $μ$ m, respectively) and vessel density (314.2 mm⁻² and 286.3 mm⁻², respectively) quantities were comparable with each other. In addition, the size and density values were consistent with previous reports.²⁴,³⁷,⁴⁹ The ratios of the MRI-derived vessel size, vessel density, and blood volume values between early and typical pseudo-normalization lesions were in a comparable range to those of LSFM-derived. However, as the thickness of the LSFM image increased, the estimated vessel density was also observed to be increasing. This is directly visible in Supplementary Figure 4, as more vessels are included with increasing thickness. Considering that MRI- and LSFM-derived vessel densities were comparable at 4 $\sim 6 μ$ m of LSFM images, the results are consistent with a previous comparison between the MRI-derived vessel density and corresponding (similar thickness) histology results.²⁴,⁵⁰,⁵¹ Such underestimation of the MRI-derived vessel density with larger thickness may be partially understood using a Monte Carlo simulation, as shown in Supplementary Figure 7. A simulated MVD was calculated using MRI experimental conditions following previous methods.²⁸ The difference between the true vessel density and simulated MVD increased as the true vessel density increased. In contrast, the simulated MVD showed a similar value (linearity) with the true vessel density when the true vessel density was under 300 mm⁻², as shown in the inset figure of Supplementary Figure 7. Further elucidation is required to rigorously verify the correlation with volumetric MRI-derived MVD and other validation methods, including 3-dimensional LSFM.

It is also noteworthy that the acquisition of LSFM images may be affected by the status of tissue clearing. Although the contralateral region showed complete tissue clearing status, ipsilateral lesions in both tMCAO rat models may not be perfectly cleared due to different tissue conditions than that of the contralateral region, which can induce bias for the absolute quantification of the vessel radius, vessel density, and blood volume. However, the ratio of vascular parameters appears to maintain close resemblance with respect to MRI measurements, which minimizes the clearing issues. In addition, the UTE-MRAs and fluorescence stereo zoom microscope image of two tMCAO rat models (without tissue clearing) showed an enlarged vessel size and reduced vessel density for early pseudo-normalization lesions, which is consistent with tissue-cleared LSFM findings, as shown in Supplementary Figures 1 and 5. Establishing an optimization method for perfectly clearing ipsilateral ischemic lesions is necessary for the absolute quantification of microvascular alterations.

In this study, along with the tissue status, morphological and functional vascular remodeling were simultaneously visualized and quantitatively analyzed as reperfusion continued from day 1 to 7 (in the subacute stage). This enabled verification of contrasting vascular remodeling tendencies between early pseudo-normalization and typical pseudo-normalization lesions. Evaluation of both morphological and functional vascular remodeling in two spatiotemporally different lesions may help better inform clinical decision making on the prognosis and treatment of ischemic stroke in the subacute stage.

Supplemental Material

sj-pdf-1-jcb-10.1177_0271678X211029197 - Supplemental material for MRI investigation of vascular remodeling for heterogeneous edema lesions in subacute ischemic stroke rat models: Correspondence between cerebral vessel structure and function

Supplemental material, sj-pdf-1-jcb-10.1177_0271678X211029197 for MRI investigation of vascular remodeling for heterogeneous edema lesions in subacute ischemic stroke rat models: Correspondence between cerebral vessel structure and function by MungSoo Kang, Seokha Jin and HyungJoon Cho in Journal of Cerebral Blood Flow & Metabolism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by grants from the National Research Foundation of Korea from the Korean government (grant Nos. 2018R1A6A1A03025810, 2018M3C7A1056887), the 2021 Joint Research Project of the Institutes of Science and Technology, the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education (grant No. 2020R1A6A3A13076670), and the Korea Healthcare Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant No. HI14C1135).

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

MSK and HJC designed the study. MSK collected the data. MSK and HJC analyzed the data. MSK and HJC wrote the manuscript. SHJ contributed to data interpretation and critical revisions of the manuscript.

Supplemental material

Supplemental material for this article is available online.

References

Mader

Schöning

Klose

, et al. Neonatal cerebral infarction diagnosed by diffusion-weighted MRI: pseudonormalization occurs early. Stroke 2002; 33: 1142–1145.

Roberts

Rowley

HA.

Diffusion weighted magnetic resonance imaging in stroke. Eur J Radiol 2003; 45: 185–194.

Schwamm

Koroshetz

Sorensen

, et al. Time course of lesion development in patients with acute stroke: serial diffusion-and hemodynamic-weighted magnetic resonance imaging. Stroke 1998; 29: 2268–2276.

Walker

Ben Salem

Lalande

, et al. Time course of NAA T2 and ADCw in ischaemic stroke patients: 1H MRS imaging and diffusion-weighted MRI. J Neurol Sci 2004; 220: 23–28.

Warach

Gaa

Siewert

, et al. Acute human stroke studied by whole brain echo planar diffusion‐weighted magnetic resonance imaging. Ann Neurol 1995; 37: 231–241.

Keep

Andjelkovic

Cytotoxic and vasogenic brain edema. In: Caplan LR, Biller J, Leary MC et al. (eds) Primer on cerebrovascular diseases. Amsterdam: Elsevier, 2017, pp.145–149.

Michinaga

Koyama

Pathogenesis of brain edema and investigation into anti-edema drugs. Int J Mol Sci 2015; 16: 9949–9975.

Liang

Bhatta

Gerzanich

, et al. Cytotoxic edema: mechanisms of pathological cell swelling. Neurosurg Focus 2007; 22: 1–9.

Rosenberg

GA.

Ischemic brain edema. Prog Cardiovasc Dis 1999; 42: 209–216.

10.

Klatzo

Pathophysiological aspects of brain edema. Acta Neuropathol 1987; 72: 236–239.

11.

Rogers

McGarry

Knight

, et al. Timing the ischaemic stroke by 1H-MRI: improved accuracy using absolute relaxation times over signal intensities. Neuroreport 2014; 25: 1180–1185.

12.

McGarry

Rogers

Knight

, et al. Stroke onset time estimation from multispectral quantitative magnetic resonance imaging in a rat model of focal permanent cerebral ischemia. Int J Stroke 2016; 11: 677–682.

13.

Jha

Cerebral edema and its management. Med J Armed Forces India 2003; 59: 326–331.

14.

Walcott

Kahle

Simard

JM.

Novel treatment targets for cerebral edema. Neurotherapeutics 2012; 9: 65–72.

15.

Rosenberg

Yang

Vasogenic edema due to tight junction disruption by matrix metalloproteinases in cerebral ischemia. Neurosurg Foc 2007; 22: 1–9.

16.

Arai

Jin

Navaratna

, et al. Brain angiogenesis in developmental and pathological processes: neurovascular injury and angiogenic recovery after stroke. FEBS J 2009; 276: 4644–4652.

17.

Krupinski

Kaluza

Kumar

, et al. Role of angiogenesis in patients with cerebral ischemic stroke. Stroke 1994; 25: 1794–1798.

18.

Bang

Saver

Kim

, et al. Collateral flow predicts response to endovascular therapy for acute ischemic stroke. Stroke 2011; 42: 693–699.

19.

Henderson

Eliasziw

Fox

, et al. Angiographically defined collateral circulation and risk of stroke in patients with severe carotid artery stenosis. Stroke 2000; 31: 128–132.

20.

Jiang

H-B

Zhang

Y-D

, et al. Collateral vessels on magnetic resonance angiography in endovascular-treated acute ischemic stroke patients associated with clinical outcomes. Oncotarget 2017; 8: 81529–81537.

21.

Ichijo

Miki

Ishibashi

, et al. Posterior cerebral artery laterality on magnetic resonance angiography predicts long-term functional outcome in Middle cerebral artery occlusion. Stroke 2013; 44: 512–515.

22.

Besselmann

Liu

Diedenhofen

, et al. MR angiographic investigation of transient focal cerebral ischemia in rat. NMR Biomed 2001; 14: 289–296.

23.

Schmidt

Villringer

, et al. Vessel size imaging reveals pathological changes of microvessel density and size in acute ischemia. J Cereb Blood Flow Metab 2011; 31: 1687–1695.

24.

Bosomtwi

Jiang

Ding

, et al. Quantitative evaluation of microvascular density after stroke in rats using MRI. J Cereb Blood Flow Metab 2008; 28: 1978–1987.

25.

Boehm‐Sturm

Farr

Adamczak

, et al. Vascular changes after stroke in the rat: a longitudinal study using optimized magnetic resonance imaging. Contrast Media Mol Imaging 2013; 8: 383–392.

26.

Moisan

Favre

Rome

, et al. Microvascular plasticity after experimental stroke: a molecular and MRI study. Cerebrovasc Dis 2014; 38: 344–353.

27.

Lin

C-Y

Chang

Cheung

W-M

, et al. Dynamic changes in vascular permeability, cerebral blood volume, vascular density, and size after transient focal cerebral ischemia in rats: evaluation with contrast-enhanced magnetic resonance imaging. J Cereb Blood Flow Metab 2008; 28: 1491–1501.

28.

Kang

Jin

Lee

, et al. Visualization of whole brain macro-and microvascular remodeling in a rat model of ischemic stroke: a pilot study. Sci Rep 2020; 10: 12. MRi

29.

Wang

Alger

Qiao

, et al. The value of arterial spin-labeled perfusion imaging in acute ischemic stroke: comparison with dynamic susceptibility contrast-enhanced MRI. Stroke 2012; 43: 1018–1024.

30.

Copen

Schaefer

MR perfusion imaging in acute ischemic stroke. Neuroimaging Clin N Am 2011; 21: 259–283.

31.

Shen

Duong

TQ.

Magnetic resonance imaging of cerebral blood flow in animal stroke models. Brain Circ 2016; 2: 20–27.

32.

Percie Du Sert

Hurst

Ahluwalia

, et al. The ARRIVE guidelines 2.0: updated guidelines for reporting animal research. J Cereb Blood Flow Metab 2020; 40: 1769–1777.

33.

Uluç

Miranpuri

Kujoth

, et al. Focal cerebral ischemia model by endovascular suture occlusion of the middle cerebral artery in the rat. J Vis Exp 2011; 48: e1978.

34.

Lee

Kang

Cho

MRI size assessment of cerebral microvasculature using diffusion-time-dependent stimulated-echo acquisition: a feasibility study in rodent. Neuroimage 2020; 215: 116784.

35.

Gage

Kipke

Shain

Whole animal perfusion fixation for rodents. J Vis Exp 2012; 65: e3564.

36.

Barbier

Lamalle

Décorps

Methodology of brain perfusion imaging. J Magn Reson Imaging 2001; 13: 496–520.

37.

Troprès

Pannetier

Grand

, et al. Imaging the microvessel caliber and density: principles and applications of microvascular MRI. Magn Reson Med 2015; 73: 325–341.

38.

Seevinck

Deddens

Dijkhuizen

RM.

Magnetic resonance imaging of brain angiogenesis after stroke. Angiogenesis 2010; 13: 101–111.

39.

Doube

Kłosowski

Arganda-Carreras

, et al. BoneJ: free and extensible bone image analysis in ImageJ. Bone 2010; 47: 1076–1079.

40.

Zhang

, et al. Cerebral hemodynamic evaluation after cerebral recanalization therapies for acute ischemic stroke. Front Neurol 2019; 10: 719.

41.

Bivard

Stanwell

Levi

, et al. Arterial spin labeling identifies tissue salvage and good clinical recovery after acute ischemic stroke. J Neuroimaging 2013; 23: 391–396.

42.

Marchal

Young

Baron

J-C.

Early postischemic hyperperfusion: pathophysiologic insights from positron emission tomography. J Cereb Blood Flow Metab 1999; 19: 467–482.

43.

Marchal

Furlan

Beaudouin

, et al. Early spontaneous hyperperfusion after stroke: a marker of favourable tissue outcome? Brain 1996; 119: 409–419.

44.

Liebeskind

Dua

, et al. Postischemic hyperperfusion on arterial spin labeled perfusion MRI is linked to hemorrhagic transformation in stroke. J Cereb Blood Flow Metab 2015; 35: 630–637.

45.

Liebeskind

, et al. ASPECTS-based reperfusion status on arterial spin labeling is associated with clinical outcome in acute ischemic stroke patients. J Cereb Blood Flow Metab 2018; 38: 382–392.

46.

Paulson

Strandgaard

Edvinsson

Cerebral autoregulation. Cerebrovasc Brain Metab Rev 1990; 2: 161–192.

47.

Hall

Reynell

Gesslein

, et al. Capillary pericytes regulate cerebral blood flow in health and disease. Nature 2014; 508: 55–60.

48.

Gjedde

Kuwabara

Hakim

AM.

Reduction of functional capillary density in human brain after stroke. J Cereb Blood Flow Metab 1990; 10: 317–326.

49.

Lemasson

Valable

Farion

, et al. In vivo imaging of vessel diameter, size, and density: a comparative study between MRI and histology. Magn Reson Med 2013; 69: 18–26.

50.

Tang

Jensen

JH.

High‐resolution MR imaging of mouse brain microvasculature using the relaxation rate shift index Q. NMR Biomed 2004; 17: 507–512.

51.

Klein

Kuschinsky

Schrock

, et al. Interdependency of local capillary density, blood flow, and metabolism in rat brains. Am J Physiol 1986; 251: H1333–H1340.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.48 MB