Evolution of cerebral ischaemia induced by thromboembolism in rats detected by early sequential MR imaging

S. Ishikawa*,1, K. Yokoyama1, T. Kuroiwa2 and K. Makita1

1Department of Anesthesiology and Critical Care Medicine, and 2Department of Neuropathology, Medical Research Institute, Tokyo Medical and Dental University, School of Medicine, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan*Corresponding author

Accepted for publication: March 26, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Thromboembolic stroke appears to evolve in patients in a very complicated manner. The present study investigated the evolution of thromboembolic stroke in rats (n=9) using a 4.7-T MR imager. Under isoflurane anaesthesia, the rats received homologous blood clots into the right internal carotid artery. After thromboembolic stroke, lesion volume, which was defined and calculated, based on apparent diffusion coefficient maps, tended to increase gradually over the 6 h magnetic resonance imaging study. The largest percentage change in lesion volume was found at the early stage (40–100 min) of thromboembolism, and showed significant correlation with total percentage change in lesion volume (41.6 (SD 32.8%)) (r=0.77, P<0.05). In conclusion, marked enlargement or diminution of lesion volume may be observed at the early stage of thromboembolism. Thromboembolic stroke, which can be partly salvaged at the early stage, may likely evolve to a lesser extent thereafter.

Br J Anaesth 2001; 87: 469–76

Keywords: complications, thromboembolism; rat; measurement techniques, magnetic resonance imaging; measurement techniques, diffusion-weighted imaging


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Perioperative stroke is uncommon, occurring in 0.08–0.2% of general surgery patients.1 2 Patients who undergo cardiac, neurologic, and vascular procedures have a higher perioperative stroke rate of 2.2–5.2%.35 Although these events are uncommon, perioperative stroke can be associated with a mortality rate of up to 26%.2

Cerebral ischaemia induced by thromboembolism remains a crucial problem during the perioperative period. It may occur silently under general anaesthesia, and may cause prolonged unconsciousness, coma, or death as a result of neurological damage. One of the first steps in an attempt to prevent or attenuate this neurological damage in patients, especially those with risk factors for thromboembolism (i.e. a previous history of transient ischaemic attack, heart disease, peripheral vascular disease),6 is to evaluate the effects on thromboembolism of neuroprotective agents or anaesthetic agents that provide some neuroprotection. However, how thromboembolic stroke evolves in humans and animals is not fully understood.

Thromboembolic stroke in patients appears to evolve in a very complicated manner. One study of patients who had suffered an acute ischaemic stroke has shown that an increase in ischaemic lesion volume occurs during the first 53 h after the stroke.7 In 12 of these 28 patients, the initial lesion volume increased by 20% or more.7 On the other hand, arterial cerebral occlusion leading to acute ischaemic stroke may disappear as a natural process of thrombus disintegration and endogenous thrombolysis. Jorgensen and colleagues reported that the incidence of spontaneous reperfusion was 77% in patients with cortical infarcts.8 To determine the evolution of thromboembolic stroke, animal experiments involving sequential measurement of ischaemic lesion volume with relatively short intervals appear necessary. However, conventional structural imaging modalities (computed tomography, T1- or T2-weighted magnetic resonance imaging (MRI)) do not reliably depict evolving ischaemic lesions at the early stage of stroke.

Diffusion-weighted imaging (DWI) is a relatively new MRI technique that is sensitive to the microscopic motion of water molecules. The signal intensity in ischaemic brain regions on DWI appears high because the diffusion coefficient of water is reduced, presumably because of the formation of cytotoxic oedema.9 10 This phenomenon can be quantified by the calculation of apparent diffusion coefficient (ADC) images.11 12 The ADC may be calculated on a pixel-by-pixel basis from at least two DWIs with different diffusion weightings. This raises the possibility that the extent and spatial evolution of thromboembolic stroke can be evaluated using an ADC map.

The rat model of cerebral infarction was established by Kudo and colleagues.13 This animal model of thromboembolism creates cerebral infarction predominantly in the territory of the middle cerebral artery, with a low mortality rate and excellent reproducibility.13 To establish a remote-controlled rat thromboembolic stroke from outside the magnet, we have modified this model and observed sequential changes in the hemispheric ischaemic lesion area caused by thromboembolism.

The objective of the present study was to determine the evolution of thromboembolic stroke in rats under general anaesthesia using ADC maps. We focused on the effect of the change in lesion volume at the early stage of stroke on total evolution of thromboembolic stroke, because recent therapy in acute stroke is based on reperfusion of the ischaemic legion leading to salvage of the ischaemic penumbra, smaller infarct size, and improved outcome.14 15


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
This project was approved by, and followed the guidelines published by, the Animal Care and Use Committee of Tokyo Medical and Dental University.

Clot formation
Nine adult male Sprague–Dawley rats weighing between 330 and 400 g were used. Anaesthesia was induced with 4% isoflurane in a Plexiglas chamber. The concentration of isoflurane was then reduced to 1.5%, 1% lidocaine was locally injected for analgesia, and 0.25 ml of blood drawn from a left femoral vein and stored at room temperature for 48 h for clot formation. The clot was fragmented by extruding it twice through a 27-gauge needle. The clot fragments were then diluted with saline solution to 0.3 ml and prepared for injection.

Animal preparation
The animals were again anaesthetized with 1.5% isoflurane, and 1% lidocaine was locally administered. Under an operating microscope, the bifurcation of the right common carotid artery was exposed through a midline neck incision. The external carotid artery was ligated just above its origin. A polyethylene catheter (PE-50) was then introduced into the right common carotid artery and advanced into the internal carotid artery. Two PE-50 catheters were introduced into a right femoral vein and a right femoral artery for drug infusion, arterial pressure monitoring, and arterial blood sampling, respectively. Physiologic saline solution at 4 ml kg–1 h–1 was administered i.v. throughout the study. After a tracheostomy, each rat was ventilated with 40% oxygen and the respiratory rate controlled to maintain PaCO2 at approximately 40 mm Hg.

MRI
MRI was performed using a 4.7-T experimental imager/spectrometer system (Unity INOVA, Varian, Palo Alto, CA, USA) with a 330-mm horizontal bore magnet equipped with shielded gradients (maximal strength, 65 mT m–1) and an 80-mm-ID quadrature detection coil. The slice plane was standardized by first obtaining preview spin-echo images. Spin-echo DWI (TR ms/TE ms=2000/64, 40-mm field of view, 128x128 matrix, 2-mm section thickness) consisted of nine coronal slices covering most of the rat brain. Diffusion weighting was applied along all three axes (b=0 and 1100 s mm–2) and a single DWI required an acquisition time of approximately 18 min. DWI with a b value of 0 s mm–2 corresponds to a T2-weighted image (T2WI).

Pairs of DWIs with two b values were used to generate maps of the ADC. ADC maps were calculated using the standard equation:16

ADC=ln(S0/S1)/(b1 b0)

where S0 and S1 are the signals of the two DWI scans and b0 and b1 are 0 and 1100 s mm–2.

Experimental protocol
After placing the animals supine in the probe, anaesthesia was maintained with 1.5% isoflurane in 40% oxygen. The rat’s body temperature was monitored with a rectal probe, and maintained at 37.5±0.5°C using a water-circulating heating pad during the MRI study. After animal positioning and shimming, a set of baseline DWIs was acquired. Homologous blood clots were introduced into the right internal carotid artery via the catheter. Sets of DWIs were acquired at 60-min intervals for 300 min beginning 40 min after clot injection (T40, T100, T160, T220, T280, T340). Arterial blood samples were collected at baseline, and at T120 and T240, and analysed on a laboratory blood gas analyser (STAT profile 5TM gas analyzer, NOVA Biomedical, USA). If mean arterial pressure (MAP) was reduced to less than 80 mm Hg, dopamine was administered i.v. to maintain the MAP over 80 mm Hg.

Definition of lesion volume
The ischaemic lesion was defined arbitrarily as the pixels in which the ADC decreased to less than 70% of the mean ADC of the ipsilateral hemisphere at baseline. Lesion area on an ADC map was calculated by multiplying the number of these pixels by the area of one pixel (approximately 0.1 mm2). The lesion volume of ischaemic injury was determined by integration of the lesion areas for the nine slices. Two coronal slices which included the optic chiasm (slice A) and the area 4 mm caudal to it (slice B) were considered to represent the territory of the middle cerebral artery, and consequently the ischaemic lesion on slices A and B were divided by the hemispheric area to obtain the percentage lesion area, and results then compared with histopathology as described below. Sequential changes in the lesion volume on ADC maps were recorded.

Percentage change in the lesion volume (%{Delta}LV) was defined as the percentage difference between the lesion volumes at Tt and Tt+60 (t=40, 100, 160, 220, 280) to calculate the percentage increase or decrease during each MRI acquisition interval. The absolute values of %{Delta}LV were compared at different times to evaluate when the lesion volume changed (increased or decreased) markedly. Total percentage change in lesion volume (total %{Delta}LV) was defined to calculate percentage increase or decrease in lesion volume during the whole MRI study after thromboembolism. %{Delta}LV and total %{Delta}LV were defined and calculated as follows.

1. As DWIs were acquired at 60-min intervals, the %{Delta}LV between Tt and Tt+60 was defined as: %{Delta}LV (Ttt+60) (%) = [[lesion volume (Tt+60) – lesion volume (Tt)]/lesion volume (Tt)]x100.

2. Total %{Delta}LV was calculated as: Total %{Delta}LV (%) = [[lesion volume (T340) – lesion volume (T40)]/lesion volume (T40)]x100.

Development of lesion area in specific regions
To analyse the development of an ischaemic lesion in some specific regions, cortex and caudate putamen on slice A and cortex and basal ganglia on slice B were selected. Then the lesion in these regions was subdivided into two parts according to their ADC values: severe lesion corresponds to ADC<400x10–6 mm2 s–1 and moderate lesion corresponds to the ischaemic lesion whose ADC value is equal to or more than 400x10–6 mm2 s–1 but less than 500x10–6 mm2 s–1. Sequential changes in lesion area were recorded and analysed.

Histopathology
After MRI studies, the animals were killed with an overdose of i.v. pentobarbital, and the brain was removed and sectioned coronally at 2-mm intervals. The sections most closely corresponding to slices A and B were selected and stained with a 2% solution of triphenyltetrazolium chloride (TTC) to identify infarcted tissue, and then photographed. Using a computer imaging system, the infarcted areas were divided by the ipsilateral hemispheric area to obtain the percentage hemispheric infarcted area.

Data analysis
All data are expressed as mean (SD). MAP, arterial blood gas data, plasma glucose concentration, rectal temperature, lesion volume (area) on ADC maps, and the absolute value of %{Delta}LV at different times were analysed using repeated measures analysis of variance followed by Tukeys multiple comparison test. The relationship between %{Delta}LV during each MRI acquisition interval and the total %{Delta}LV was analysed using linear fitting. The relationship between lesion volume on the ADC map and the time after clot injection was analysed for each rat using linear fitting. The percentage lesion area based on the ADC map at T340 was compared with the TTC stained percentage hemispheric infarcted area using correlation analysis for slices A and B. Statistical significance was set at the P<0.05 level.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
No significant differences in MAP, heart rate, arterial blood gas, plasma glucose concentration, or rectal temperature at different time points throughout the study were observed (Table 1). Dopamine infusion was required in one rat to maintain MAP over 80 mm Hg.


View this table:
[in this window]
[in a new window]
 
Table 1 Haemodynamic indices, arterial blood gas, plasma glucose concentration, and rectal temperature at baseline, 100, 220, and 340 min after clot injection. Results are mean (SD). There were no significant differences in any index between any time points. MAP=mean arterial pressure; HR=heart rate
 
ADC maps, DWIs, and T2WIs on slices A and B before and after thromboembolism in a representative rat are shown in Figures 1 and 2, respectively. DWI revealed a region of increased signal intensity in the territories of the right middle cerebral artery in all rats after clot injection. A clear reduction in the ADC value was evident in these regions after embolization. Lesion area was not evident on T2WI at T100 but apparent at T220 in both slices.



View larger version (127K):
[in this window]
[in a new window]
 
Fig 1 ADC maps, diffusion- and T2-weighted images (DWI and T2WI) on slice A at baseline and 100, 220, and 340 min after thromboembolism.

 


View larger version (139K):
[in this window]
[in a new window]
 
Fig 2 ADC maps, diffusion- and T2-weighted images (DWI and T2WI) on slice B at baseline and 100, 220, and 340 min after thromboembolism.

 
Mean lesion volumes did not differ significantly at the various time points. Lesion volume in individual rats showed some fluctuation during the MRI study (Figure 3). Lesion volume on the ADC map was larger at T340 than at T40 in eight of nine rats. There was a significant linear relationship between lesion volume and the time after thromboembolism in five rats, and the mean (SD) slope and intercept were 0.43 (0.31) mm3 min–1 and 243.9 (96.1) mm3, respectively.



View larger version (20K):
[in this window]
[in a new window]
 
Fig 3 Sequential change in lesion volume. Mean lesion volumes (dashed line) did not differ significantly at the various time points. The lesion volume in individual rats showed some fluctuation during the MRI study (thick lines).

 
%{Delta}LV during each MRI acquisition interval is shown in Figure 4. The absolute value of %{Delta}LV (T40–100) (22.7 (17.2)%) was two to three times larger than that at T100–160, T160–220, T220–280, and T280–340. It was significantly larger (P<0.05) than that at T160–220 (7.9 (7.2)%) and T280–340 (8.1 (8.5)%). Mean (SD) total %{Delta}LV was 41.6 (32.8)%. Correlation analysis revealed a significant linear correlation between %{Delta}LV (T40–100) and total %{Delta}LV (r=0.77, P<0.05) (Figure 5). No significant correlation was observed between %{Delta}LV at other time intervals and total %{Delta}LV.



View larger version (14K):
[in this window]
[in a new window]
 
Fig 4 Percentage change in lesion volume (%{Delta}LV) at each MRI acquisition interval. The absolute value of %{Delta}LV at 40–100 min after thomboembolism was two or three times larger than at other intervals.

 


View larger version (14K):
[in this window]
[in a new window]
 
Fig 5 Correlation analysis between percentage change in lesion volume at 40–100 min after thromboembolism (%{Delta}LV (T40–100)) and total percentage change in lesion volume (total %{Delta}LV). There was a significant relationship (r=0.77, P<0.05).

 
Severe ischaemic lesions (ADC<400x10–6 mm2 s–1) significantly increased with time in the cortex on both slices. In the caudate putamen on slice A and basal ganglia on slice B, mean severe ischaemic lesions doubled during MRI study, but there was no significant differences at the various time points (Figure 6). There was also no significant differences in moderate ischaemic lesions (400x10–6 mm2 s–1<=ADC<500x10–6 mm2 s–1) at different time points in any selected region (Figure 7).



View larger version (16K):
[in this window]
[in a new window]
 
Fig 6 Sequential changes in mean severe ischaemic lesions in the cortex (•), and caudate putamen ({circ}) on slice A and cortex ({blacksquare}) and basal ganglia ({square}) on slice B. Severe ischaemic lesions significantly increased with time in the cortex on both slices. SD are not shown for clarity. *P<0.05, **P<0.01 vs 40 min.

 


View larger version (15K):
[in this window]
[in a new window]
 
Fig 7 Sequential changes in mean moderate ischaemic lesions in the cortex (•), and caudate putamen ({circ}) on slice A and cortex ({blacksquare}) and basal ganglia ({square}) on slice B. There was no significant differences in moderate ischaemic lesions at different time points in any selected region. SD are not shown for clarity.

 
Mean (SD) percentage hemispheric infarcted area determined by TTC was 62.9 (17.8) and 56.8 (17.2)% on slices A and B, respectively. Postmortem TTC studies showed that all the animals had an infarction that correlated well to the hyperintense area on DWI and the area with a low ADC value on the ADC map. Significant correlation between the percentage hemispheric infarcted area and the hemispheric lesion area on the final ADC map for slices A and B was observed (Figure 8).



View larger version (12K):
[in this window]
[in a new window]
 
Fig 8 Percentage hemispheric infarcted area (%, x) and percentage lesion area on ADC map at 340 min after thromboembolism (%, y). There was a significant correlation on slices A (•, solid line) and B ({blacksquare}, dashed line). (A) y=1.23x + 8.57 (r=0.88, P<0.01). (B) y=1.00x + 3.33 (r=0.80, P<0.01).

 

    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We have assessed the evolution of thromboembolic stroke in rats using ADC maps. The lesion area was defined based on ADC maps for nine coronal slices and integrated to calculate lesion volume. Mean lesion volume tended to increase with time although there was no significant difference at various time points and in individual rats there was some fluctuation. Lesion volume showed a maximal change at T40–100, and a significant correlation was observed between the percentage change in lesion volume during that period (%{Delta}LV (T40–100)) and the total change in lesion volume (total %{Delta}LV).

Diffusion-weighted MR imaging has been well established as a reliable non-invasive method for early detection of cerebral ischaemic stroke.17 18 The principle of diffusion measurement is that the spatial location of each water molecule is tagged such that any net movement of the water molecule over time results in a signal loss or darkening of the images. In the present study, this tagging was achieved by a pair of very strong pulsed magnetic field gradients (diffusion gradients) which were added to a spin-echo sequence. The first of these diffusion gradients causes the spinning protons to fall out of phase with one another, and the second gradient rephases the protons completely with one another only if there has been no movement of the water molecules between the application of the two diffusion gradients. When there is movement of water, the protons are not brought back completely into phase by the second diffusion gradient and the signal attenuation will occur on DWIs. DWI can detect abnormalities within min after the onset of ischaemia in animal models of stroke.1921 before lesions are apparent on conventional images. The high signal intensity on the DWIs after clot injection may be explained as an ischaemic change as a result of thromboembolism. The infarct lesions revealed by TTC staining may also suggest that the hyperintense areas on DWIs are ischaemic because of decreased blood flow, which is consistent with the histopathological findings of Kudo and colleagues.13

After injection of the clot, the ADC values were reduced in regions where the signal intensities were increased on the DWIs. Although the precise mechanism underlying the reduction in ADCs in ischaemic regions is unclear, this phenomenon is considered to represent the onset of intracellular water accumulation or cytotoxic oedema.10 22 23 Tissue adenosine triphosphate (ATP) reserves are depleted after the onset of ischaemia, and the subsequent failure of the Na+/K+-ATPase pump causes water protons and ions to migrate from the faster-diffusing, extracellular space into the slower-diffusing, more restricted intracellular compartment, resulting in an ADC decrease.

Thromboembolic stroke in patients evolves in a very complicated manner. For example, the ischaemic lesion may enlarge with time7 and result in large cerebral infarction. On the other hand, spontaneous reperfusion may occur8 and ischaemic penumbra may be salvaged. Many studies have shown a qualitative enlargement of thromboembolic stroke in rats but not quantitative changes in lesion volume in individual rats.2429 Thus, we designed the present study to examine how lesion volume after thromboembolism evolves quantitatively using ADC maps. Lesion volumes were measured sequentially at relatively short intervals (60 min). The precise mechanisms responsible for the change in lesion volume after thromboembolism are not fully understood. We assume many factors are involved, as described below.

First, mean percentage lesion volume tended to increase with time after thromboembolism in the present study; however, the increase was not statistically significant. This enlargement of lesion volume may be most likely explained by an ischaemic penumbra. The lesion observed on the first ADC map after the thromboembolic stroke may be regarded as the central most severely ischaemic region (‘core’30). The lesion area that was not visible on the ADC map on the early stage of the thromboembolism but was detected later during the MRI study may represent ‘penumbra’.30 31 Although cerebral perfusion was not evaluated in the present study, the most likely scenario is that cerebral blood flow in the ‘penumbra’ reduced slightly and the ADC decreased little or did not change32 on the first ADC map after thromboembolism. This hypoperfused tissue may be above the threshold for infarction, at least initially;33 however, it may be irreversibly injured on the time course of the MRI study.

Development of severe ischaemic lesions in the cortex on slices A and B may support our assumption that the enlargement of lesion volume may be most likely explained by an ischaemic penumbra. The increase in severe ischaemic lesion area may be explained that the regions with higher ADC value (>=400x10–6 mm2 s–1) became injured during the time course of the experiment. The area of moderate lesion was unchanged because the higher ADC regions (>=500x10–6 mm2 s–1) may have been injured and showed a decrease in their ADC value to 400–500x10–6 mm2 s–1.

Second, as shown in Figure 3, the lesion volume in individual rats showed some fluctuation during the MRI study. Transient reduction in lesion volume (transient recoveries of ADCs) in such animals may be, in part, explained by recanalization and subsequent reperfusion (spontaneous reperfusion8 34), a passage of emboli or a development of adequate collateral circulation. Disappearance of arterial cerebral occlusion is often seen clinically as a natural process of thrombus disintegration and endogenous thrombolysis. Although it is impossible to prove in the present study without measurement of cerebral perfusion, the proportional relationship between ADC and cerebral perfusion32 may indirectly support our assumption.

Third, pseudonormalization of the ADC may be another factor that affected the time course of lesion volume. In rodent stroke models, ADC begins to normalize at 24–48 h. This normalization does not imply tissue recovery but the beginning of tissue necrosis (pseudonormalization).35 36 As ADC is increased by vasogenic oedema,22 pseudonormalization is considered to be attributed to a balance of cytotoxic and vasogenic oedema. Lesion area was evident on T2WI at T220 in the present study, suggesting that vasogenic oedema occurred at that time. Lesion volume based on ADC maps may have been affected by pseudonormalization of ADC during the latter half of the time course of the MRI study; however, we assume the extent of underestimation of lesion volume because of pseudonormalization, if it occurred, may be small as the TTC stained infarcted area correlated well with lesion volume on the final ADC map.

The fact that the absolute value of %{Delta}LV was largest at T40–100 may imply that lesion volume markedly increases or decreases at the early stage of thromboembolism. Irrespective of the mechanisms responsible for the sequential change observed in lesion volume, our data demonstrate that total %{Delta}LV shows significant correlation to %{Delta}LV at this period (T40–100) but not thereafter. This suggests that enlargement or diminution of lesion volume during this period plays an important role in the total evolution of thromboembolic stroke in this model. Our results also suggest that only a small increase or even a decrease in lesion volume may be detected throughout the whole period of thromboembolism in rats in which the ischaemic lesion is partly salvaged at the early stage of thromboembolic stroke.

There is no absolute threshold of ADC decrease that predicts if a lesion is irreversible; maximal decreases in the ADC area are even reversible for short periods.37 This is why the threshold ADC values should be set arbitrarily. The mean (SD) ADC value on slices A and B at baseline were 718 (40)x10–6 and 703 (44)x10–6 mm2 s–1, respectively. We have set the threshold at 70% of the baseline, which corresponded to approximately 500x10–6 mm2 s–1, by modifying the study by Busch and colleagues.38 Morikawa and colleagues reported on the adverse effect of hyperglycaemia on transient focal brain ischaemia using the threshold ADC value of 500x10–6 mm2 s–1.39 One may expect that –2 SD from the mean ADC at baseline would be better as the threshold ADC. The mean –2 SD ADC on slices A and B correspond to 583x10–6 and 539x10–6 mm2 s–1, respectively in the present study, which would have estimated the lesion volume (area) a little larger than our analysis.

In conclusion, the largest percentage change in lesion volume was found at the early stage of thromboembolism, and showed significant correlation with the total percentage change in lesion volume. Marked enlargement or diminution of the lesion volume may be observed at the early stage of thromboembolism. Thromboembolic stroke, which is partly salvaged at the early stage, may likely evolve to a lesser extent thereafter.


    Acknowledgements
 
The authors greatly appreciate Masato Ueki, MD, for his technical support and advice. This study was partly supported by a Grant-in-Aid (C) (No. 09671542) from the Ministry of Education, Science and Culture, Japan. The present study was partly presented at 1999 American Society of Anesthesiologists annual meeting.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Larsen SF, Zaric D, Boysen G. Postoperative cerebrovascular accidents in general surgery. Acta Anaesthesiol Scand 1988; 32: 698–701[ISI][Medline]

2 Parikh S, Cohen JR. Perioperative stroke after general surgical procedures. NY State J Med 1993; 93: 162–5

3 Breuer AC, Furlan AJ, Hanson MR, et al. Central nervous system complications of coronary artery bypass graft surgery: prospective analysis of 421 patients. Stroke 1983; 14: 682–7[Abstract]

4 Tuman KJ, McCarthy RJ, Najafi H, Ivankovich AD. Differential effects of advanced age on neurologic and cardiac risks of coronary artery operations. J Thorac cardiovasc Surg 1992; 104: 1510–7[Abstract]

5 Turnipseed WD, Berkoff HA, Belzer FO. Postoperative stroke in cardiac and peripheral vascular disease. Ann Surg 1980; 192: 365–8[ISI][Medline]

6 Caplan LR. Diagnosis and treatment of ischemic stroke. JAMA 1991; 266: 2413–8[ISI][Medline]

7 Baird AE, Benfield A, Schlaug G, et al. Enlargement of human cerebral ischemic lesion volumes measured by diffusion-weighted magnetic resonance imaging. Ann Neurol 1997; 41: 581–9[ISI][Medline]

8 Jorgensen HS, Sperling B, Nakayama H, Raaschou HO, Olsen TS. Spontaneous reperfusion of cerebral infarcts in patients with acute stroke. Arch Neurol 1994; 51: 865–73[Abstract]

9 Fisher M, Sotak CH, Minematsu K, Li L. New magnetic resonance techniques for evaluating cerebrovascular disease. Ann Neurol 1992; 32: 115–22[ISI][Medline]

10 Benveniste H, Hedlund LW, Johnson A. Mechanism of detection of acute cerebral ischemia in rats by diffusion-weighted magnetic resonance microscopy. Stroke 1992; 23: 746–54[Abstract]

11 Bihan DL, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986; 161: 401–7[Abstract]

12 Bihan DL, Moonen CTW, van Zijl PCM, Pekar J, DesPres D. Measuring random microscopic motion of water in tissues with MR imaging: a cat brain study. J Comput Assist Tomogr 1991; 15: 19–25[ISI][Medline]

13 Kudo M, Aoyama A, Ichimori S, Fukunaga N. An animal model of cerebral infarction: homologous blood clot emboli in rats. Stroke 1982; 13: 505–8[Abstract]

14 Baird AE, Donnan GA, Austin MC, Fitt GJ, Davis SM, McKay WJ. Reperfusion after thrombolytic therapy in ischemic stroke measured by single-photon emission computed tomography. Stroke 1994; 25: 79–85[Abstract]

15 Clark WM, Albers GW, Madden KP, Hamilton S. The rtPA (alteplase) 0- to 6-hour acute stroke trial, part A (A0276g): results of a double-blind, placebo-controlled, multicenter study. Stroke 2000; 31: 811–6[Abstract/Free Full Text]

16 Le Bihan D, Breton E, Lallemand D, et al. Separation of diffusion and perfusion in intravoxel incoherent motion MR imaging. Radiology 1988; 168: 497–505[Abstract]

17 Rordorf G, Koroshetz WJ, Copen WA, et al. Regional ischemia and ischemic injury in patients with acute middle cerebral artery stroke as defined by early diffusion-weighted and perfusion-weighted MRI. Stroke 1998; 29: 939–43[Abstract/Free Full Text]

18 van Everdingen KJ, van der Grond J, Kappelle LJ, Ramos LMP, Mali WPTM. Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke 1998; 29: 1783–90[Abstract/Free Full Text]

19 Minematsu K, Fisher M, Li L, et al. Effects of a novel NMDA antagonist on experimental stroke rapidly and quantitatively assessed by diffusion-weighted MRI. Neurology 1993; 43: 397–403[Abstract]

20 Minematsu K, Fisher M, Li L, Sotak CH. Diffusion and perfusion magnetic resonance imaging studies to evaluate a non competitive N-methyl-D-aspartate antagonist and reperfusion in experimental stroke in rats. Stroke 1993; 24: 2074–81[Abstract]

21 Lo EH, Matsumoto K, Pierce AR, Garrido L, Luttinger D. Pharmacologic reversal of acute changes in diffusion-weighted magnetic resonance imaging in focal cerebral ischemia. J Cereb Blood Flow Metab 1994; 14: 597–603[ISI][Medline]

22 Ebisu T, Naruse S, Horikawa Y, et al. Discrimination between different types of white matter edema with diffusion-weighted MR imaging. J Magn Reson Imaging 1993; 3: 863–8[Medline]

23 Sevick RJ, Kanda F, Mintorovitch J, et al. Cytotoxic brain edema: assessment with diffusion-weighted MR imaging. Radiology 1992; 185: 687–90[Abstract]

24 Busch E, Kruger K, Fritze K, Allegrini PR, Hoehn-Berlage M, Hossmann KA. Blood-brain barrier disturbances after rt-PA treatment of thromboembolic stroke in the rat. Acta Neurochir 1997; 70(Suppl.): 206–8

25 Busch E, Kruger K, Peter R, et al. Reperfusion after thrombolytic therapy of embolic stroke in the rat: magnetic resonance and biochemical imaging. J Cereb Blood Flow Metab 1998; 18: 407–18[ISI][Medline]

26 Jiang Q, Zhang RL, Zhang ZG, Ewing JR, Divine GW, Chopp M. Diffusion-, T2-, and perfusion-weighted nuclear magnetic resonance imaging of middle cerebral artery embolic stroke and recombinant tissue plasminogen activator intervention in the rat. J Cereb Blood Flow Metab 1998; 18: 758–67[ISI][Medline]

27 Jiang Q, Zhang RL, Zhang ZG, et al. Magnetic resonance imaging indexes of therapeutic efficacy of recombinant tissue plasminogen activator treatment of rat at 1 and 4 h after embolic stroke. J Cereb Blood Flow Metab 2000; 20: 21–7[ISI][Medline]

28 Zhang R, Chopp M, Zhang ZG, Jiang Q, Ewing JR. A rat model of focal embolic cerebral ischemia. Brain Res 1997; 766: 83–92[ISI][Medline]

29 Zhang Z, Zhang RL, Jiang Q, Raman SBK, Cantwell L, Chopp M. A new rat model of thrombotic focal cerebral ischemia. J Cereb Blood Flow Metab 1997; 17: 123–35[ISI][Medline]

30 Astrup J, Symon L, Siesjö BK. Thresholds in cerebral ischemia – the ischemic penumbra. Stroke 1981; 12: 723–5[ISI][Medline]

31 Memezawa H, Smith M-L, Siesjo BK. Penumbral tissues salvaged by reperfusion following middle cerebral artery occlusion in rats. Stroke 1992; 23: 552–9[Abstract]

32 Pierce AR, Lo EH, Mandeville JB, Gonzalez RG, Rosen BR, Wolf GL. MRI measurements of water diffusion and cerebral perfusion: their relationship in a rat model of focal cerebral ischemia. J Cereb Blood Flow Metab 1997; 17: 183–90[ISI][Medline]

33 Jones TH, Morawetz RB, Crowell RM, et al. Thresholds of focal cerebral ischemia in awake monkeys. J Neurosurg 1981; 54: 773–82[ISI][Medline]

34 Baird AE, Austin MC, Mckay WJ, Donnan GA. Changes in cerebral tissue perfusion during the first 48 h of ischaemic stroke: relation to clinical outcome. J Neurol Neurosurg Psychiatry 1996; 61: 26–9[Abstract]

35 Welch KM, Windham J, Knight RA, et al. A model to predict the histopathology of human stroke using diffusion and T2-weighted magnetic resonance imaging. Stroke 1995; 26: 1983–9[Abstract/Free Full Text]

36 Garcia JH, Liu KF, Ho KL. Neuronal necrosis after middle cerebral artery occlusion in Wistar rats progresses at different time intervals in the caudoputamen and the cortex. Stroke 1995; 26: 636–42[Abstract/Free Full Text]

37 Davis D, Ulatowski J, Eleff S, et al. Rapid monitoring of changes in water diffusion coefficients during reversible ischemia in cat and rat brain. Magn Reson Med 1994; 31: 454–60

38 Busch E, Beaulieu C, de Crespigny A, Moseley ME. Diffusion MR imaging during acute subarachnoid hemorrhage in rats. Stroke 1998; 29: 2155–61[Abstract/Free Full Text]

39 Morikawa S, Inubushi T, Ishii H, Nakasu Y. Effects of blood sugar level on rat transient focal brain ischemia consecutively observed by diffusion-weighted EPI and 1H echo planar spectroscopic imaging. Magn Reson Med 1999; 42: 895–902[ISI][Medline]