Functional, perfusion and diffusion MRI of acute focal ... · Combined functional, perfusion and diffusion magnetic resonance imaging (MRI) with a temporal resolution of 30mins was

Functional, perfusion and diffusion MRI of acutefocal ischemic brain injury

Qiang Shen1, Hongxia Ren1, Haiying Cheng1, Marc Fisher2, Timothy Q Duong1

1Department of Neurology, Yerkes Primate Research Center, Emory University, Atlanta, Georgia, USA;2Department of Neurology, University of Massachusetts Medical School, Worcester, Massachusetts, USA

Combined functional, perfusion and diffusion magnetic resonance imaging (MRI) with a temporalresolution of 30 mins was performed on permanent and transient focal ischemic brain injury in ratsduring the acute phase. The apparent diffusion coefficient (ADC), baseline cerebral blood flow(CBF), and functional MRI (fMRI) blood-oxygen-level-dependent (BOLD), CBF, and CMRO2

responses associated with CO2 challenge and forepaw stimulation were measured. An automatedcluster analysis of ADC and CBF data was used to track the spatial and temporal progression ofdifferent tissue types (e.g., normal, ‘at risk,’ and ischemic core) on a pixel-by-pixel basis. Withpermanent ischemia (n¼ 11), forepaw stimulation fMRI response in the primary somatosensorycortices was lost, although vascular coupling (CO2 response) was intact in some animals. Controlexperiments in which the right common carotid artery was ligated without causing a stroke (n¼ 8)showed that the delayed transit time had negligible effect on the fMRI responses in the primarysomatosensory cortices. With temporary (15-mins, n¼ 8) ischemia, transient CBF and/or ADCdeclines were observed after reperfusion. However, no T2 or TTC lesions were observed at 24 hexcept in two animals, which showed very small subcortical lesions. Vascular coupling and forepawfMRI response also remained intact. Finally, comparison of the relative and absolute fMRI signalchanges suggest caution when interpreting percent changes in disease states in which the baselinesignals are physiologically altered; quantitative CBF fMRI are more appropriate measures. Thisapproach provides valuable information regarding ischemic tissue viability, vascular coupling, andfunctional integrity associated with ischemic injury and could have potential clinical applications.Journal of Cerebral Blood Flow & Metabolism (2005) 25, 1265–1279. doi:10.1038/sj.jcbfm.9600132; published online27 April 2005

Keywords: BOLD; CBF; CMRO2; DWI; fMRI; ISODATA; penumbra; perfusion–diffusion mismatch; PWI

Introduction

Diffusion-weighted (DWI) and perfusion-weighted(PWI) magnetic resonance imaging (MRI) techniquesoffer the potential to noninvasively classify ‘normal’,‘at-risk,’ and ‘ischemic’ tissues associated withstroke during the acute phase. Brain tissues withperfusion deficits below a critical threshold (Hos-smann, 1994) experience metabolic energy failure,membrane depolarization, and subsequent cellularswelling. These changes precipitate a reduction inthe apparent diffusion coefficient (ADC) of brainwater and are manifested as hyperintense regions on

DWI (Moseley et al, 1990). During the acute phase,the DWI abnormality is initially smaller relative tothe area of perfusion deficit. As ischemia evolves,most of this DWI abnormality expands and even-tually coincides with the abnormal PWI area. Thedifference in the abnormal region defined by thePWI and DWI in acute stroke is commonly referredto as the ‘perfusion–diffusion’ mismatch. It has beensuggested that the ‘perfusion–diffusion’ mismatchis potentially salvageable and approximates the‘ischemic penumbra’ (Astrup et al, 1981).

With the advent of functional MRI (fMRI), itbecomes possible to assess the integrity of vascularcoupling and neural function of ‘at-risk’ and‘salvaged’ tissues, as defined by DWI and PWI.Anatomically defined ‘at-risk’ and salvaged tissuesmay or may not show functional deficit and thusfMRI adds a unique and important measure to strokeimaging. The most widely used fMRI technique isbased on the blood-oxygen-level-dependent (BOLD)signal (Ogawa et al, 1990) and the cerebral bloodflow (CBF) signal. The BOLD contrast originates

Received 19 January 2005; revised 9 March 2005; accepted 14March 2005; published online 27 April 2005

Correspondence: Dr TQ Duong, Department of Neurology,Yerkes National Primate Research Center, Emory University, 954Gatewood Road NE, Atlanta, GA 30329, USA.E-mail: [email protected]

This work was supported in part by the NIH (NINDS, R01-

NS45879) and the American Heart Association (SDG-0430020N).

Journal of Cerebral Blood Flow & Metabolism (2005) 25, 1265–1279& 2005 ISCBFM All rights reserved 0271-678X/05 $30.00

www.jcbfm.com

from intravoxel magnetic field inhomogeneity in-duced by paramagnetic deoxyhemoglobin in the redblood cells. Changes in regional deoxyhemoglobincontent can be visualized in susceptibility-sensi-tized (i.e., T2

*-weighted) BOLD images. The BOLDfMRI is based on a principle discovered over 100years ago (Roy and Sherrington, 1890) that neuronalactivity is intricately coupled to CBF. When a task isperformed, regional CBF increases disproportion-ally, which can be measured using the arterial-spin labeling MRI technique. The CBF increaseovercompensates the stimulus-evoked increase inoxygen consumption needed to fuel the elevatedneural activity, resulting in a regional reductionin deoxyhemoglobin concentration. Thus, the CBFand BOLD fMRI signals increase during stimulus-evoked changes in neural activity relative to basalconditions, making it possible to dynamically andnoninvasively map changes in brain functions.

Furthermore, Davis et al introduced an eloquentformalism based on the BOLD biophysical model(Ogawa et al, 1993) to estimate stimulus-evokedCMRO2 changes based on BOLD and CBF (orcerebral blood volume (CBV)) measurements. Theadvantage of Davis’ formalism over existing formal-isms is that there are no a priori assumptionsregarding resting capillary or venous oxygen satura-tion, blood volume fraction, blood flow, and meta-bolic rate of oxygen. All these parameters and otherphysiologic quantities are lumped into the constantM, which can be measured on a pixel-by-pixelbasis using a calibration procedure involving CO2

challenge. Although this CMRO2 model has yetto be validated, especially in ischemic injury,estimates of stimulus-evoked CMRO2 changes instroke could yield interesting and valuable informa-tion regarding tissue metabolism.

The use of fMRI to evaluate functional statusrequires preservation of the hemodynamic andmetabolic coupling. Carbon dioxide, a potent vaso-dilator, plays an important role in modulating bloodflow associated with increased neuronal activity.The hypercapnia model in which CO2 is added tothe breathing gas has been widely used to examinethe vascular coupling (Weiss et al, 1983; Ueki et al,1988; Kim and Ugurbil, 1997; Davis et al, 1998; Hogeet al, 1999; Corfield et al, 2001; Duong et al, 2001).One major advantage of hypercapnic challenge isthat it increases blood flow, and thus the BOLDsignal, of the entire brain without changing neuralactivity or metabolism, thereby making possible theevaluation of vascular coupling without the addedeffects of metabolic changes. Previous studies onblood flow regulation after prolonged ischemiasuggest that CO2 reactivity is severely disturbedafter ischemic injury (Schmidt-Kastner et al, 1986;Ueki et al, 1988).

While a few fMRI studies on stroke patients havebeen reported (Pineiro et al, 2002; Guadagno et al,2003; Binkofski and Seitz, 2004), similar studies inanimal stroke models have been limited. Animal

models where focal ischemia can be reproduciblystudied under controlled conditions are importantfor systemically evaluating the integrity of vascularcoupling and functional status of ‘at risk’ andsalvaged tissues defined by perfusion and diffusionimaging. Functional MRI using an exogenous CBVcontrast agent (MION) and BOLD contrast haverecently been applied to study stroke rats during thechronic phase to evaluate functional reorganization(Dijkhuizen et al, 2001, 2003). Functional MRIassessment of the dynamically evolving ischemictissue during acute stroke, however, have not beenreported, likely because of the stringent need forhigh spatiotemporal resolution and the lack of asuitable animal model. Most of stroke studies in ratsused chloral hydrate as the anesthetic (Li et al, 1999,2000; Shen et al, 2003) whereas essentially allforepaw-stimulation fMRI studies in normal ratsuse a-chloralose (Mandeville et al, 1998; Silva et al,1999; Duong et al, 2000). a-Chloralose became apopular anesthetic for fMRI studies in rats because itcan induce nonresponsiveness without depressingthe central nervous system to the same extent asmany other common anesthetics. However, animalsrecovering from a-chloralose anesthesia exhibitatypical behavior (Lees, 1972) and thus a-chloraloseis only approved for terminal studies. In addition,fMRI studies under a-chloralose typically requiremechanical ventilation with repeated blood-gassampling to optimize the fMRI responses and it isrelatively difficult to maintain a constant level ofanesthesia with injectable anesthetics in general(Austin et al, 2005) for eliciting consistent fMRIresponses from the onset (i.e., during acute ische-mia) and throughout the studies. More recently, arobust forepaw stimulation model in rats wasshowed under isoflurane anesthesia and sponta-neously breathing conditions where the animals canmaintain their own physiology throughout theexperiment without the need for repeated blood-gas sampling (Liu et al, 2004a). This model could bewell suited for repeated fMRI studies of acute andchronic stroke in a longitudinal fashion.

In this study, we developed and applied acombined diffusion, perfusion, and functional MRIprotocol to study focal cerebral ischemia during theacute phase with a 30-mins temporal resolution andreasonable spatial resolution. Quantitative perfusionand diffusion imaging was performed and an im-proved ISODATA clustering technique (Shen et al,2004) was used to classify normal, ischemic core,and ‘diffusion–perfusion mismatch’ tissues basedon ADC and CBF characteristics. Blood-oxygen-level-dependent, CBF, and CMRO2 fMRI associatedwith hypercapnic challenge and forepaw stimula-tion were used to evaluate the integrity of vascularcoupling and neural function, respectively, asischemia progressed during the acute phase. Blood-oxygen-level-dependent, CBF, and CMRO2 fMRIresponses in different ISODATA-derived ischemictissue types and in the forepaw somatosensory

Functional, perfusion and diffusion MRI of acute strokeQ Shen et al

1266

Journal of Cerebral Blood Flow & Metabolism (2005) 25, 1265–1279

cortices of the right ischemic hemisphere (RH) wereanalyzed and compared with the homologousregions in the normal left hemisphere (LH). Thisapproach was tested in the permanent and transient(15-mins) acute focal brain ischemia in rats underisoflurane, breathing spontaneously.

Materials and methods

Animal Preparation

Twenty seven male Sprague–Dawley rats (300 to 350 g,Taconic Farms, NY, USA) were anesthetized with 2%isoflurane in air during surgery. Three groups of animalswere studied: (i) permanent focal brain ischemia of the RHwas induced with intraluminal middle cerebral arteryocclusion (MCAO) (n¼ 11, Group I), (ii) transient (15-mins) focal brain ischemia (n¼ 8, Group II) was inducedusing the same method and reperfusion was performedoutside the magnet, and (iii) the right common carotidartery (CCA) was ligated but without causing a stroke toevaluate the possible effect of delayed transit time on thefMRI responses (n¼ 8, Group III). Measurements weremade before and during CCA ligation in the same animalswhere the CCA occlusion was achieved remotely whilethe animals were in the magnet. The left femoral arterywas catheterized for blood-gas sampling and physiologicmonitoring. Once the rats were secured in an MR-compatible rat stereotaxic headset, anesthesia wasreduced to 1.1% to 1.2% isoflurane. Rats breathedspontaneously without mechanical ventilation. Rectaltemperature was maintained at 37.01C70.51C. Heart rate(HR) and mean arterial blood pressure (MABP) viathe arterial line were recorded continuously onto a PCvia the Biopac system (Santa Barbara, CA, USA). Respira-tion rate (RR) was derived from the slow modulations ontop of the cardiac waveforms. Blood gas was typicallysampled once during a break between imaging trials. Allrecorded physiologic parameters were within normalphysiologic ranges. Magnetic resonance imaging datawere acquired at 30, 90, and 180 mins and again at 24 hafter ischemia.

Hypercapnic Challenge and Forepaw Stimulation

Hypercapnic challenges used a premixed gas of 10% CO2

with 21% O2 and balance N2. Forepaw somatosensorystimulation used the previously optimized parametersunder identical isoflurane anesthetic condition in normalanimals (Liu et al, 2004a): 6 mA current with 0.3 ms pulseduration at 3 Hz. These stimulation parameters did notcause an increase in MABP. Needle electrodes wereinserted under the skin of the two forepaws before surgery.The electrodes were connected in series and the twoforepaws were stimulated simultaneously. During eachtime point (30-mins imaging block), one trial of hyper-capnic challenge and two trials of forepaw stimulationwere presented with a 4-mins break (without imaging)between trials. Each trial consisted of 4 mins of dataacquired during baseline and 2 mins of data acquiredduring hypercapnic challenge or forepaw stimulation.

MR Experiments

Magnetic resonance imaging experiments were performedon a 4.7-T/40-cm magnet, a Biospec Bruker console(Billerica, MA, USA), and a 20-G/cm gradient insert(ID¼ 12 cm, 120-ms rise time). A surface coil (2.3-cm ID)was used for brain imaging and a neck coil (Silva et al,1999; Duong et al, 2000) for perfusion labeling. Coil-to-coilelectromagnetic interaction was actively decoupled. Acomplete imaging block lasted 30 mins where ADC, basalCBF, hypercapnic, and forepaw-stimulation fMRI wereacquired.

ADCav was obtained by averaging three ADC mapswith diffusion-sensitive gradients separately appliedalong the x, y or z direction. Single shot, echo-planarimages (EPI) were acquired with matrix¼ 64� 64, spectralwidth¼ 200 kHz, TR¼ 2 secs (901 flip angle), TE¼ 37.5 ms,b¼ 4 and 1,170 secs/mm2, D¼ 24 ms, d¼ 4.75 ms, FOV¼2.56 cm� 2.56 cm, eight 1.5-mm slices, and 16 averages(total time B2.5 mins).

Combined CBF and BOLD measurements were madeusing the continuous arterial spin-labeling technique withsingle-shot, gradient-echo, EPI acquisition. Paired imageswere acquired alternately—one with arterial spin labelingand the other without (control). MR parameters were asfollows: data matrix¼ 64� 64, FOV¼ 2.56 cm� 2.56 cm,eight 1.5-mm slices, TE¼ 20 ms, and TR¼ 2 secs (901 flipangle). Continuous arterial spin labeling employed a 1.78-s square radiofrequency pulse to the labeling coil in thepresence of 1.0 G/cm gradient along the flow directionsuch that the condition of adiabatic inversion wassatisfied. The sign of the frequency offset was switchedfor control (nonlabeled) images. For each set of CBF andBOLD measurements, 60 pairs of images (4 mins) wereacquired during baseline and 30 pairs (2 mins) duringhypercapnic challenge or forepaw stimulation.

Data Analysis

Data analysis used codes written in Matlab (MathWorksInc., Natick, MA, USA) and the STIMULATE (Universityof Minnesota) software. Coregistration of images obtainedduring the acute phase and at 24 h used an in-housesoftware, which involved both manual and automaticalignment without spatial interpolation, as describedpreviously (Liu et al, 2004a).

Calculation of basal ADC and CBF maps: Apparentdiffusion coefficient maps with intensity in unit of mm2/sec were calculated pixel-by-pixel by using (Stejskal andTanner, 1965), ADC¼�ln (S1/S0)/(b1–b0) where bi¼ g2 Gi

2

d2(D–d/3), where ln is the natural logarithm, S0 and S1 arethe signal intensities obtained with b0 and b1, respectively.The b-value is proportional to the gradient strength (G),magnetogyric ratio (g), duration of each gradient pulse (d),and the time (D) between applications of the two gradientpulses. Apparent diffusion coefficient maps were calcu-lated at each time point.

Cerebral blood flow images (SCBF) with intensity inunits of mL/g/min were calculated (Silva et al, 1999;


1267


Duong et al, 2000) pixel-by-pixel using, SCBF¼l/T1(SC�SL)/(SLþ (2a�1) SC), where SC and SL are signalintensities of the control and labeled images, respectively,and l is the partition coefficient and a is the labelingefficiency. Basal CBF images were obtained from thebaseline period (without stimulation) of all three trials ofthe fMRI measurements within each time point.

ISODATA analysis of diffusion and perfusionimages: An improved ISODATA clustering algorithmwas used to statistically resolve and identify tissues as‘normal’, ‘mismatch’ and ‘ischemic core’ based on tissueADC and CBF characteristics as described elsewhere(Shen et al, 2004). Only five anterior slices were analyzedto avoid susceptibility distortion around the ear canalsassociated with the posterior slices. ROIs of the wholebrain were carefully drawn to avoid the edge of the brain-skull interface based on CBF maps with reference to ADCmaps. Pixels with ADC higher than 1.0� 10�3 mm2/sec(i.e., CSF pixels) and the corpus callosum were excludedfrom analysis. Clustering for ADC and baseline CBF wasperformed individually at each time point for each animal.Three clusters were resolved: (1) pixels with normal ADCand baseline CBF were assigned to be ‘normal’ tissue,(2) pixels with low ADC and baseline CBF were assignedto be ‘ischemic core’, and (3) pixels with normal ADC butlow baseline CBF were assigned to be ‘perfusion–diffusionmismatch’.

Functional MRI responses in different ISODATA clus-ters were analyzed. Scatterplots of CBF versus ADC andBOLD versus ADC under baseline and stimulated (CO2 orforepaw) conditions were generated at each time point.Different ISODATA clusters were color-coded on thescatterplots.

Evolution of ‘mismatch’ pixels: The temporal andspatial evolution of the ‘mismatch’ pixels, defined at30 mins after occlusion, was evaluated as they migratedto different clusters. Apparent diffusion coefficient, CBF,and BOLD under baseline and stimulated (CO2 or forepaw)conditions were analyzed for the pixels that subsequentlymigrated into the normal zone, core zone, or remainedin the mismatch zone at 180-mins after ischemia. Forpermanent occlusion, ischemia stopped evolving 180 minsafter occlusion, which was taken as the imaging endpoint as shown previously (Shen et al, 2003; Meng et al,2004).

Calculation of M and CMRO2 changes: CMRO2 wascalculated using (Davis et al, 1998),

DBOLD

BOLD0¼ M 1 � CMRO2

CMRO2 0j

� �b CBF

CBF0

� �a�b !

ð1Þ

where M is the proportionality constant and parameterswith subscript zero indicate baseline values. Grubb’sfactor (a) of 0.38 (Grubb et al, 1974; Mandeville et al,1999) and b of 1.5 (Boxerman et al, 1995; Davis et al, 1998)were taken to be constants, reflecting the effects of bloodvolume and deoxyhemoglobin concentration on the BOLDsignal, respectively. M values were calculated from thehypercapnia (CBF and BOLD) data using the same

equation where the CMRO2/CMRO2|0 was set to unitybecause brief and mild hypercapnia does not alter CMRO2

(Kety and Schmidt, 1948; Novack et al, 1953). Stimulus-evoked CMRO2 changes from an ROI of the primarysomatosensory cortex were derived from the stimulus-evoked BOLD and CBF time courses associated withforepaw stimulation. Additional details of the modelincluding error propagation have been described else-where (Liu et al, 2004a).

Hypercapnic responses in different tissue types:Images obtained during the transition period betweenbaseline and stimulus onset (30 secs for CO2 challenge and15 secs for forepaw stimulation) were discarded. Blood-oxygen-level-dependent images were obtained from thecontrol (nonlabeled) images of the CBF measurements.Blood-oxygen-level-dependent and CBF magnitude andpercent changes relative to baselines were calculated:(1) on a pixel-by-pixel basis, (2) for the ISODATA-derivednormal, mismatch, and core clusters, and (3) for the ROI ofthe forepaw somatosensory cortices.

Forepaw-stimulation responses in the forepawcortices: Cross correlation analysis associated with theforepaw primary somatosensory cortices was performed.ROI of the normal LH forepaw cortices were drawn basedon the averaged cross correlation activation maps of alltime points with references to the rat brain atlas and MRIanatomical images to avoid bias to any particular timepoint. The forepaw ROIs on the ischemic RH wereobtained by symmetrically reflecting the LH ROIs alongthe midline to the RH. Apparent diffusion coefficient,baseline CBF, and fMRI signals in the forepaw primarysomatosensory cortex ROI were analyzed pixel-by-pixelas well as by averaging pixels within the forepawROI. Magnitude baseline CBF and CBF changes werecomputed. Baseline T2

*-weighted signal intensities(within the forepaw ROI) of the ischemic RH wasnormalized with respect to the homologous ROI of theLH for comparison.

All data values in text are expressed as mean7s.d. andin graphs as mean7s.e.m. A P-value o0.05 (t-test) wastaken to be statistically significant.

Results

Permanent Occlusion (Group I): CO2 Responses inDifferent ISODATA-Derived Clusters

Images and maps at 30 mins after ischemia of onerepresentative animal subjected to permanent focalischemia are shown in Figure 1A. Basal CBF imagesdepict a large perfusion deficit and ADC imagesdepict a smaller ADC lesion in the ischemic RH.Nine out of 11 rats showed a substantial andevolving ‘perfusion–diffusion’ mismatch and a fewof these animals showed some persistent mismatchat 180 mins. The forepaw primary somatosensorycortices typically showed an evolving a perfusion–diffusion mismatch because of its anatomical


1268


location close to the anterior communicating artery.The two remaining animals showed large infarcts atthe onset without apparent mismatch at 30 mins.ISODATA analysis yielded three distinct tissuetypes in the RH, namely, normal (blue pixels),mismatch (green pixels), and core (red pixels). The

fMRI maps showed that the hypercapnic challengeevoked a marked CBF increases in the LH, slightlyattenuated CBF increases in the RH ‘normal’ tissuesand essentially no evoked CBF increases in the RHischemic ‘core’ and ‘mismatch’ tissues. Hypercapnia-induced CBF increases were essentially absent in

CBFbaseline

ADC

ISODATA

CBFCO2 map

BOLDCO2 map

30 mins post-occlusion

ADC (x 10-3mm2/s)

∆CBFCO2 ∆BOLDCO2BOLDbaseline

Nor

mal

ized

BO

LD s

igna

ls

LH

RH

CB

F (

mL/

g/m

in)

4

3

2

1

0

-1

4

3

2

1

0

-1

2.5

2.0

1.5

1.0

0.5

0.0

2.5

2.0

1.5

1.0

0.5

0.0

0.50.40.30.20.1

0-0.1-0.2

0.50.40.30.20.1

0-0.1-0.2

4

3

2

1

0

-1

4

3

2

1

0

-1

CBFbaseline

0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0

0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0

A

B

Figure 1 Permanent occlusion—CO2 functional magnetic resonance imaging (fMRI) responses in different ISODATA-derived clusters(n¼11). (A) Representative CBF and apparent diffusion coefficient (ADC) maps, ISODATA clusters overlaid on ADC maps, DCBFCO2

and DBOLDCO2percent change maps overlaid on CBF images of a rat subjected to permanent focal ischemia at the 30-min time

point. ISODATA cluster analysis yielded ‘normal’ (blue), ‘perfusion–diffusion mismatch’ (green), and ‘ischemic core’ (red) clusters.Gray-scale bar indicates ADC ranging from 0 to 0.001 mm2/sec and CBF 0 to 3 mL/g/min. Color bar indicates DCBF ranging from10% to 400% and DBOLD 1% to 10%. (B) Scatterplots of baseline CBF versus ADC, DCBFCO2

versus ADC, normalized baseline T2*-

weighted signal intensities versus ADC, and DBOLDCO2versus ADC for the left hemisphere (LH) and the right hemisphere (RH) at

30 mins after ischemia (one animal). ISODATA-derived clusters were assigned as ‘normal’ (blue), ‘perfusion–diffusion mismatch(green) and ‘ischemic core’ clusters (red). (C) Group-average baseline CBF, DCBFCO2

, and DBOLDCO2of the normal left hemisphere

and the ISODATA-derived normal, mismatch, core clusters of the right ischemic hemisphere at three time points after ischemia. (D)Temporal evolution of the group-average baseline CBF, DCBFCO2

, and DBOLDCO2of the ‘mismatch’ before and after migration to

different zones as ischemia progressed. Mismatch at 30 mins after ischemia was used as the reference and these mismatch pixelswere tracked as they migrated to different zones at 180 mins after ischemia.


1269


regions with perfusion deficit. Interestingly,although dependent on the precise threshold used,the BOLD activated areas appeared to be larger thanthe CBF activation areas, suggesting differences inmetabolic and CBF changes associated with is-chemic injury.

The ADC and CBF values of different ISODATA-derived clusters are summarized in Table 1. LH ADCand LH basal CBF were similar at all time points.In the normal zone of the ischemic RH, ADC andCBF were normal. In the mismatch zone, ADCwas marginally reduced but CBF was significant

reduced. In the core zone, both ADC and CBF weremarkedly reduced.

Pixel-by-pixel scatterplot analysis of CBF versusADC under baseline and CO2 challenge from oneanimal is shown in Figure 1B (left panel). The LHbasal CBF was 1.270.7 mL/g/min (whole brainaverage). Hypercapnia markedly increased CBF,with 90% of all pixels showing a pixel-by-pixelDCBF increase ranging from 0.5 to 3 mL/g/min. Inthe RH, the normal, mismatch and core clusterswere derived using the ISODATA analysis and colorcoded. Essentially all ‘core’ and ‘mismatch’ pixels

0.0

0.5

1.0

1.5

2.0

2.5 1 11

22

2

3 33 444

180min90min30min 180min90min30min

baseline ∆CBFCO

2

baseline

CB

F (

mL/

g/m

in)

0

2

4

6

8

4

4

4

3

3

3

2

22

11

1

BO

LD p

erce

nt c

hang

esB

OLD

per

cent

cha

nge

1: LH 2: normal 3: mismatch 4: core

0.0

0.2

0.4

0.6

0.8

1.0

1.2

∆CBFCO2

CB

F (

mL/

g/m

in)

normal mismatch core 30min



normal mismatch core180min

0

1

2

3

4

5

6

C

D

Figure 1 (Continued)

Table 1 Group average ADC and CBF for different time point and different area (n¼11, mean7s.d.)

ADC (�10�3 mm2/sec) CBF baseline (mL/g/min)

30 min 90 min 180 min 30 min 90 min 180 min

LH 0.7570.02 0.7670.01 0.7570.01 0.8970.21 0.9770.17 0.9770.20Normal 0.7470.02 0.7470.02 0.7470.03 0.8170.20 0.8370.20 0.8170.15Mismatch 0.6970.02 0.7070.02 0.7070.03 0.1370.04 0.1070.03 0.1570.06Core 0.5170.02 0.4870.02 0.4670.03 0.1370.12 0.1570.20 0.1370.11

ADC, apparent diffusion coefficient; CBF, cerebral blood flow; LH, left hemisphere.


1270


showed no CO2-induced CBF increases. Most of theRH ‘normal’ pixels showed significant CO2-inducedCBF increases but were smaller in magnituderelative the LH.

The scatterplot analysis of BOLD versus ADC isshown on the right panel of Figure 1B. The widedistribution of the normalized baseline T2

*-weightedsignal intensities in the LH is indicative of tissueheterogeneity, tissue physiologic status (i.e., hyp-oxia), and nonuniform signal intensity profile ofthe surface coil. Nonetheless, in the LH, theCO2-induced BOLD increases were evident.Although the RH ‘core’ pixels showed lower base-line T2

*-weighted signal intensities because ofhypoxia-induced increase in deoxyhemoglobinconcentration and/or changes in blood volume,different ISODATA clusters (derived via ADC andbaseline CBF data) were not well segregated basedon T2

*-weighted signal intensities. CO2-inducedDBOLD versus ADC showed slightly better clusterseparation visually but nonetheless were signifi-cantly less resolved compared with the correspond-ing CO2-induced DCBF versus ADC. The averageBOLD fMRI responses were 3.5%, 3.3%, 1.1%, and�0.8% in the normal LH, normal RH, mismatch, andcore clusters, respectively.

The group-average hypercapnia-induced CBF andBOLD changes in the LH and in different RHISODATA clusters at different time points aresummarized in Figure 1C. For permanent occlusion,ischemia stopped evolving 180 min after occlusion(Shen et al, 2003; Meng et al, 2004), which wastaken as the imaging end point. Baseline CBF of theRH normal cluster was similar to that of the LH.However, the RH CO2-induced CBF and BOLD fMRIresponses, although robust, were attenuated relativeto the LH fMRI responses. There were trends ofimprovements in the RH fMRI responses over time.Baseline CBF of the RH mismatch and ischemic corewere markedly reduced and their CO2-induced fMRIresponses were essentially absent. BOLD responsesin the RH mismatch and ischemic core were alsomarkedly attenuated. Note that only BOLD percentchanges were plotted because baseline T2

*-weightedintensities could not be compared across differentbrain regions because of tissue heterogeneity andnonuniform intensity profile of the surface coil.

Tracking the Perfusion–Diffusion Mismatch

To better understand the tissue at risk, mismatchpixels were tracked as ischemia progressed. Themismatch pixels at 30 min were taken as referencesand individual pixels were dynamically tracked interms of their basal CBF, CBF, and BOLD fMRIresponses as they migrated to different zones at180 min (Figure 1D). Majority of the mismatch pixelsmigrated to the ‘ischemic core’ at 180 min afterischemia, consistent with the definition of ischemicpenumbra. Pixels that migrated to the core zone at

180 min after ischemia showed ADC decreases (0.6770.03 at 30 min to 0.4870.02 at 180 min); those thatremained in the mismatch zone or those thatmigrated to the normal zone showed no ADCdecreases relative to the normal LH. As expected,pixels migrating to the ‘normal’ zone showed aslightly higher baseline CBF, DCBF, and DBOLDresponses at 30 min and significantly higher base-line CBF, DCBF, and BOLD responses at 180 min. Incontrast, pixels that remained in the mismatch zoneor migrated to the ischemic ‘core’ zone showedlower baseline CBF, DCBF, and DBOLD responses atboth 30 and 180 min. Although the magnitudes ofthe RH CO2-induced DCBF were markedly reduced,the CO2-induced CBF percent changes, notably,were significant and comparable with those of thenormal LH because of division by the small RHbaseline values under ischemia. This observationsuggests caution in interpreting relative percentchanges in disease state.

Permanent Occlusion (Group I): CO2 and ForepawResponses in the Primary Somatosensory Cortices

In addition to analyzing the CO2 response profiles ofdifferent ISODATA-derived tissue types, functionalintegrity associated with forepaw stimulation inischemic brain injury was also evaluated. CBFversus ADC and BOLD versus ADC scatterplots ofpixels within the forepaw somatosensory corticesassociated with hypercapnic challenge and forepawstimulation are showed in Figure 2A. In the normalLH, CO2 challenge and forepaw stimulation evokedsignificant changes in the CBF and BOLD fMRIresponses (red pixels) relative to the baseline (greenpixels), with the stimulus-evoked CBF increasesshowing significantly larger dynamic ranges. In theischemic RH, the center of mass of pixels associatedwith CO2 challenge showed small but observableCBF and BOLD increases. Stimulus-evoked changesin the center of mass of these pixels in the ischemicRH were, however, visually negligible.

Hypercapnic responses within the primary soma-tosensory cortices are summarized in Figure 2B (toppanel). In the normal LH across all time points,baseline CBF in the forepaw somatosensory cortexwas B1.05 mL/g/min, CO2 inhalation increased CBFto B3 mL/g/min, and forepaw stimulation increasedCBF to B1.4 mL/g/min. In the ischemic RH, baselineCBF in the forepaw somatosensory cortex was0.25 mL/g/min, CO2 challenge induced a small CBFincrease of o0.2 mL/g/min. However, if the datawere expressed as percent changes, CO2-inducedCBF increases were 50 to 70%, again suggestingcaution in interpreting relative percent changesin physiologically perturbed state. Baseline T2

*-weighted signal intensities of the RH were normal-ized with respective to those of the normal LHforepaw cortex for comparison. Assuming that thesignal drop off within the small forepaw primary


1271


somatosensory area due to the surface coil applica-tion was negligible, baseline normalized T2

*-weighted signal intensities between the LH and RH

could be compared. Under basal condition withoutstimulation, normalized baseline T2

*-weighted imageintensities in the ischemic RH were slightly reduced

CB

F (

mL/

g/m

in)

Left hemisphere Right hemisphere Left hemisphere Right hemisphereCO

2

Forepaw

CO2

Forepaw

CB

F (

mL/

g/m

in)

Nor

mal

ized

BO

LD

4.5

3.5

2.5

1.5

0.5

-0.5

2.0

1.5

1.0

0.5

0.0

2.0

1.5

1.0

0.5

0.0

2.0

1.5

1.0

0.5

0.0

Nor

mal

ized

BO

LD

2.0

1.5

1.0

0.5

0.0

2.5

2.0

1.5

1.0

0.5

0.0

-0.5

2.5

2.0

1.5

1.0

0.5

0.0

-0.5

0.90

0.95

1.00

1.05

1.10

Nor

mal

ized

BO

LD s

igna

ls

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

CB

F (

mL/

g/m

in)

0.90

0.92

0.94

0.96

0.98

1.00

1.02

1.04

1.06

Nor

mal

ized

BO

LD s

igna

ls

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

RHLH

baseline∆CBF

baseline

∆CBF

baseline

∆BOLD

baseline

∆BOLD

CB

F (

mL/

g/m

in)

0.2 0.4 0.6 0.8 1.0

1.00.2 0.4 0.6 0.8 1.00.2 0.4 0.6 0.8 1.00.2 0.4 0.6 0.8 1.00.2 0.4 0.6 0.8

0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0

ADC (x 10-3mm2/s)

30 90 180 30 90 180 (min)RHLH

30 90 180 30 90 180 (min)

RHLH30 90 180 30 90 180 (min)

RHLH30 90 180 30 90 180 (min)

A

B

4.5

3.5

2.5

1.5

0.5

-0.5

Figure 2 Permanent occlusion—CO2 and forepaw functional magnetic resonance imaging (fMRI) responses in the primarysomatosensory cortices (n¼11). (A) Group-average scatterplots of baseline CBF versus apparent diffusion coefficient (ADC),DCBFCO2

versus ADC, normalized baseline T2*-weighted signal intensities versus ADC, and DBOLDCO2

versus ADC for the lefthemisphere and the right hemisphere (RH) at 180 mins after ischemia. Data were obtained from the forepaw somatosensorycortices. The green pixels are baseline conditions and red pixels are stimulus-evoked changes. (B) Group-average fMRI responses toCO2 challenge and forepaw stimulation in the forepaw somatosensory cortices. Baseline CBF and DCBF fMRI responses, normalizedT2

*-weighted signal intensities and DBOLD fMRI responses are shown for different time points after ischemia. The inset showsrepresentative CBF fMRI activation maps of forepaw stimulation.


1272


by 2.5% as expected because of hypoxia and/orblood volume changes, as reported previously(Roussel et al, 1995; van der Toorn et al, 1995). Inthe normal LH, CO2 challenge increased BOLD byB7.4% whereas, in the ischemic RH, CO2 challengeincreased BOLD by 2.5%. This is in marked contrastto the magnitude stimulus-evoked CBF responsesthat was negligible in the ischemic RH. Thisdifference between BOLD and CBF responses asso-ciated with forepaw stimulation is consistent withthat associated with CO2 challenge (Figure 1B).

The forepaw stimulation responses within theprimary somatosensory cortices are shown in Figure2B (bottom panel). Baseline CBF, baseline T2

*-weighted signal intensities, and fMRI responsesdid not vary substantially across different timepoints after ischemia. In the normal LH, forepawevoked robust BOLD and CBF responses. In theischemic RH, forepaw stimulation evoked a negli-gible CBF increase. However, BOLD increase wasnot negligible. This difference between BOLD andCBF responses associated with forepaw stimulationis consistent with that associated with CO2 chal-lenge (Figures 1A and 2B top panel), suggestinguncoupling of metabolic activity and blood flow.

To better understand the potential uncoupling ofmetabolism and CBF, stimulus-evoked CMRO2

changes associated with forepaw stimulation inpermanent ischemic brain injury were also esti-mated. Table 2 summarizes the M values andstimulus-evoked CMRO2 changes in the primarysomatosensory cortices in the LH and RH at differenttime points after occlusion. M is the proportionalityconstant in the CMRO2 model and it reflects thehemodynamic coupling. M value in the normal LHwas B10% across different time points whereas thatin the ischemic RH was reduced by half, ca. B5%.Forepaw-stimulation-induced CMRO2 increase wasB10% in the LH and B3% in the RH.

Transient (15-mins) Occlusion (Group II)

Images and maps at 30 mins after ischemia ofone representative animal subjected to transient

(15-mins) focal ischemia are shown in Figure 3A.All eight animals exhibited regions of hypoperfu-sion at the first imaging time point (B30 mins) butonly one animal showed detectable ADC reductionat 30 mins. Six animals showed heterogeneous andmild hypoperfusion at 180 mins, but no ADC lesionswere observed. Six animals showed no T2 or TTClesions at 24 h after ischemia and two showed verysmall T2 and TTC abnormality (43 mm3) at 24 h inthe subcortical region. CBF images at 24 h werenormal with no apparent hypo- or hyper-perfusion.Hypercapnic challenge evoked marked CBF andBOLD increases in both LH and RH and there wereno statistical differences in CO2-induced fMRIresponses between the LH and RH at all time points(P40.05). Basal CBF and basal T2

*-weighted signalintensities were not statistically different betweenthe LH and RH somatosensory cortices (P40.05).Similarly, there were also no statistical differencesin the stimulus-evoked CBF and BOLD fMRIresponses between the LH and RH somatosensorycortices (P40.05).

Stimulus-evoked CMRO2 changes and M values inthe LH and RH primary somatosensory cortices aftertransient ischemic brain injury were estimated(Table 3). M values were not statistically differentbetween LH and RH primary somatosensory cortices(P40.05) and they ranged from 7.7% to 9.4% acrossdifferent time points. Forepaw-stimulation-inducedCMRO2 increases in the LH and RH were also notstatistically different between LH and RH primarysomatosensory cortices (P40.05) and they rangedfrom 13% to 21% across different time points.

Effects of Mean Transit Time on fMRI Responses:CCA Occlusion Without Causing Stroke (Group III)

Attenuation of the fMRI responses could be becauseof ischemia and/or delayed mean transit timeassociated with occlusion of major arteries. Toinvestigate the potential effects of delayed meantransit time on the fMRI responses, control fMRIexperiments were performed in which the commoncarotid artery was clipped (occluded) remotely butwithout causing a stroke as confirmed by ADCmeasurements (data not shown). Functional MRIresponses obtained from the primary somatosensorycortices are summarized in Figure 4. Although boththe baseline and stimulus-evoked CBF under theclip condition were marginally lower relative to theno-clip condition, there were no statistical differ-ences between clip and no-clip data (P40.05).Subcortical structures, however, showed signifi-cantly delayed MTT effect. The effective whole-brain CBF values were reduced by 20%.

Discussion

The major results of this study are: (1) Combineddiffusion, perfusion, and fMRI of ischemic brain

Table 2 Permanent occlusion: group-average percent M andCMRO2 from the primary forepaw somatosensory cortices(n¼11, mean7s.d.)

LH RH

M (%)30 mins 11.072.0 7.372.190 mins 10.972.1 7.173.1180 mins 10.672.1 7.072.4

DCMRO2 (%)30 mins 9.673.5 3.472.390 mins 10.973.4 3.172.0180 mins 10.573.1 4.372.6

RH, right hemisphere; LH, left hemisphere.


1273


ADC

CBF

CO2

Forepaw

CO2

Forepaw

30 min

24 hr

30 min

24 hr

30 min

24 hr

30 min

24 hr

0.80

0.85

0.90

0.95

1.00

1.05

1.10

1.15

RHLH

Nor

mal

ized

BO

LD s

igna

ls

30m 90m 180m

90m 180m 90m 180m

24h 30m 90m 180m 24h 30m 90m 180m 24h 30m90m180m 24h

RHLH30m 90m 180m 24h 30m90m180m 24h

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

RHLH

CB

F (

mL/

g/m

in)

30m 24h 30m 24h0.0

0.5

1.0

1.5

2.0

2.5 Basal CBF

∆CBF

Basal CBF

∆CBF

Basal BOLD

∆BOLD

Basal BOLD

∆BOLD

RHLH

CB

F (

mL/

g/m

in)

0.90

0.95

1.00

1.05

1.10

Nor

mal

ized

BO

LD s

igna

ls

A

B

Figure 3 Transient occlusion–CO2 and forepaw functional magnetic resonance imaging (fMRI) responses in the primarysomatosensory cortices (n¼8). (A) Apparent diffusion coefficient (ADC) maps, baseline CBF, DCBFCO2

, and forepaw stimulationDCBF of a rat subjected to transient ischemia. Gray-scale bar indicates ADC ranging from 0 to 0.001 mm2/sec and CBF 0 to 3 mL/g/min. Color bar indicates DCBF ranging from 10% to 400% and DBOLD 1% to 10%. (B) Group-average baseline CBF, DCBFCO2

, andDBOLDCO2

of the normal left hemisphere (LH) and the right hemisphere (Rh) ISODATA-derived normal, mismatch, core clusters ofthe right ischemic hemisphere.


1274


injury during the acute phase was performed witha temporal resolution of 30 mins. (2) AutomatedISODATA cluster analysis of MR tissue character-istics (ADC, CBF, and fMRI responses) was im-plemented, making it possible to spatially andtemporally track different tissue types on pixel-by-pixel basis, providing valuable informationregarding tissue viability, vascular coupling, andfunctional integrity associated with the dynamicevolution of ischemic brain injury. (3) Permanent

cerebral ischemia resulted in a complete loss ofneuronal activity in the primary somatosensorycortices in all animals whereas vascular couplingwas intact in some animals. Loss of vascularcoupling and/or functional integrity preceded ADClesion and were temporally associated with theperfusion decline. Thus, fMRI could potentiallyserve as an earlier and complementary indictor forrisk of ischemic brain injury. (4) Transient (15-mins)cerebral ischemia exhibited transient CBF and/orADC decline. However, no permanent injury wasobserved in essentially all animals at 3 and 24 h.Vascular coupling and fMRI responses in theforepaw cortices were also normal. (5) The effect ofdelayed transit time on the fMRI responses in theprimary somatosensory cortices due to occlusion ofmajor arteries in this stroke model was negligible.(6) Cerebral blood flow fMRI responses appeared tohave larger dynamic ranges than the BOLD res-ponses and pixel-by-pixel CBF changes were moreapparent, suggesting a potentially better mappingsignal. (7) Finally, caution must be exercised wheninterpreting relative fMRI signal changes in physio-logically perturbed states in which the baselinesignals are altered, resulting in erroneous percentchanges. Measurements of magnitude changes(such as quantitative CBF changes) might benecessary to provide accurate measures of increasedneural activity under physiologically perturbedconditions.

Table 3 Transient (15 mins) occlusion: group-average M andCMRO2 from the primary forepaw somatosensory cortices(n¼8, mean7s.d.)

LH RH

M (%)30 mins 9.270.8 9.470.490 mins 9.470.6 8.670.5180 mins 8.270.3 7.770.424 h 8.872.1 9.171.6

DCMRO2 (%)30 mins 1373 137390 mins 1974 1873180 mins 1673 177624 h 21710 23713

RH, right hemisphere; LH, left hemisphere.

0.00.20.40.60.81.01.21.41.61.8

CB

F (

mL/

g/m

in)

baseline forepaw stim

0.96

0.97

0.98

0.99

1.00

1.01

1.02

1.03 baseline forepaw stim

BO

LD s

igna

l (no

rmal

ized

)

no-clip clip no-clip clip

no-clip clip no-clip clip

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2

CB

F (

mL/

g/m

in)

baseline CO2

0.96

0.98

1.00

1.02

1.04

1.06

1.08 baseline CO2

BO

LD s

igna

l (no

rmal

ized

)

CO2

Forepaw

A B

Figure 4 Common carotid artery occlusion without causing stroke—CO2 and forepaw functional magnetic resonance imaging (fMRI)responses in the primary forepaw somatosensory cortices (n¼8). (A) Group-average baseline CBF and DCBF associated with CO2

challenge and forepaw stimulation. (B) Group-average normalized BOLD and DBOLD associated with CO2 challenge and forepawstimulation.


1275


Errors in CBF Measurements

Although consistent with many established techni-ques, the accuracy of this CBF technique could besubject to errors from magnetization-transfer (Silvaet al, 1995), transit-time (Calamante et al, 1996;Zhou et al, 2001), and water-exchange (Silva et al,1997; Zhou et al, 2001; Parkes and Tofts, 2002)effects. The magnetization-transfer effect was not anissue with the actively decoupled two-coil system(Duong et al, 2000; Silva et al, 2000). Transit-timeeffect in small animals and water-exchange effect aresmall (Silva et al, 1997; Zhou et al, 2001; Parkes andTofts, 2002) and unlikely to alter the conclusions ofthis study. Ischemia-induced changes in transittime, however, could result in significant errors andremains to be vigorously validated. In the controlexperiments in which the CCA was occluded, it wasestimated that the upper limit of CBF error wasB20% (whole brain). Despite the potential errorassociated with delayed transit time in stroke,quantitative CBF fMRI arguably is a better measureconsidering the alternative fMRI percent changes,which incorrectly reflect neural activity becauseof the marked changes in baseline values underischemic conditions.

Advantages and Disadvantages of ForepawStimulation fMRI of Acute Stroke UnderIsoflurane Anesthesia

Forepaw stimulation is a good model for assessingthe functional integrity in the rat stroke modelherein because the primary somatosensory corticesgenerally experience an evolving perfusion–diffu-sion mismatch, most of which will inevitablybecome infarcted if left untreated. Although widelyused in fMRI studies of normal animals, a-chloraloseis less suited for fMRI studies of both acute andchronic ischemic brain injury. In this study, wechose to use isoflurane under spontaneously breath-ing condition which has been previously optimizedto yield robust responses (Liu et al, 2004a). As thisstudy shows, fMRI studies could be performedduring the acute and chronic phase of ischemicbrain injury.

The disadvantage is that isoflurane is a potentvasodilator and a strong neural suppressant, whichcould attenuate the hemodynamic and neuralresponses, respectively. Nonetheless, robust fMRIresponses could be detected as shown herein andpreviously in normal animals (Liu et al, 2004a;Sicard and Duong, 2005). Isoflurane has also beensuggested to have some neuroprotective effects(Mortier et al, 2000; Elsersy et al, 2004). Ourunpublished data showed that the infarct sizesunder isoflurane with permanent focal ischemiawere not statistically different from that underchloral hydrate, whereas the infarct size underisoflurane with transient focal ischemia wassmaller, depending on experimental conditions.

Nonetheless, the choice of isoflurane anestheticshould not affect the overall conclusion of thisstudy, namely, that these MRI techniques and theanimal model offer a unique means to dynamicallystudy the vascular coupling and functional integrityof different ischemic tissue types as a function ofischemic evolution in both acute and chronic phase.

Magnitude versus Percent Changes

A striking observation is that although the magni-tudes of the CO2-induced DCBF in the ischemic RHwere markedly reduced, the CO2-induced CBFpercent changes were substantial and comparablewith those of the normal LH (Figure 2B). This isbecause of the division by small RH baseline valuesassociated with ischemia. Comparison of the mag-nitude and percent change suggests that cautionsmust be exercised when interpreting fMRI databased on fMRI percent changes in disease stateswhere the baseline physiology is markedly altered.We further suggest that measurements of magnitudechanges are needed for fMRI studies underperturbed conditions. Similar observations werereported associated with fMRI associated withsystemic cocaine administration which caused sub-stantial transient changes in mean arterial bloodpressure, heart rate and respiration (Liu et al,2004b). These physiologic effects on stimulus-evoked fMRI responses had been systemicallyevaluated in another study (Sicard and Duong,2005) in which the relative and quantitative CBF,normalized (with respect to a fixed baseline) BOLDand CMRO2 fMRI changes associated with forepawstimulation in normal animals were measuredwhere the baseline conditions were physiologicallymodulated by changing the inhaled CO2 or O2 gasconcentrations. It was found that quantitative CBF,normalized CMRO2 and BOLD signals are accurateindicators of increased neural activity under mildphysiologic perturbation whereas relative fMRIsignal changes are not.

Differences in Hypercapnic and Functional Responses

In some animals, there was persistent mismatch at180 mins and the RH forepaw somatosensory cortexdid not show ADC lesions at 180 mins and 24 h. Itwas also noted that the RH forepaw somatosensorycortices did respond to CO2 challenge but did notrespond to forepaw stimulation (Figures 1A and 2B)in some animals. One potential explanation is thatthe RH forepaw somatosensory cortex was func-tional silence because of the systemic ischemicshock although the RH forepaw somatosensorycortex was not injured. The other potential explana-tion is that the functional connection from whichthe cortex receives its inputs (i.e., thalamus) wassevered, although both vascular coupling and func-tions of the RH forepaw somatosensory cortex were


1276


intact. As such, no forepaw-stimulation fMRI re-sponses in the RH somatosensory cortices weredetected despite their normal ADC and CO2 fMRIresponses. This notion is supported by the extendedsubcortical injury but remains to be validated.

Differences between CBF and BOLD fMRI Responses

In the ischemic RH, the CO2 BOLD activated areaswere larger than the CO2 CBF activated areas (Figure1A) and the CO2 and forepaw BOLD responses wereattenuated less than the CO2 and forepaw CBFresponses (Figure 2B). This observation suggestsuncoupling of metabolism, CBV and/or CBF asso-ciated with ischemia. For example, CMRO2 could bereduced, leading to a larger BOLD increase. There isevident that basal CBV elevates in ischemic injury(Roussel et al, 1995; van der Toorn et al, 1995),which could modulate the BOLD percent changes. Itshould be noted that activated areas in Figure 1B aredependent on the precise thresholds used. Becausethe BOLD and CBF contrast-to-noise ratios weredifferent, different thresholds were used whichcould potentially lead to bias. In Figure 1B, thresh-olds in the CO2 challenge studies were judicially setfor display such that the central ischemic coreshowed no activated pixels. In Figure 2B, however,ROI analysis of the primary forepaw somatosensorycortices was performed without using thresholdsand thus there was no bias. In brief, the differencesbetween CBF and BOLD fMRI responses are likelygenuine, indicative of metabolic and hemodynamicuncoupling associated with ischemic brain injury.

Imaging CMRO2 in Acute Stroke

To further investigate the apparent metabolic andhemodynamic uncoupling, we attempted, as a firststep, to estimate CMRO2 changes associated withischemic brain injury. The M values in the normalLH (10% and 8% in the permanent and transientischemic model, respectively) were higher than that(ca., 5%) reported previously with a shorter TE of15 ms in normal rats under essentially identicalpreparation (Liu et al, 2004a). This difference wasexpected because a shorter TE yields a smaller Mvalue (see equation(1)). Nonetheless, this differencedoes not affect the CMRO2 determination as long asthe same MR parameters were used in both the Mand CMRO2 measurements in a single setting.Forepaw-stimulation-induced CMRO2 increases inthe normal LH (10% and 17% in the permanent andtransient ischemic model, respectively) were smal-ler than that of the 24%76% increase in normalanimals reported previously (Liu et al, 2004a). Wehave no explanation for this discrepancy. A poten-tial explanation for the apparent discrepancy instimulation-induced CMRO2 changes between theprevious and current study could be that the strokesurgery per se could cause a reduced stimulus-

evoked CMRO2 change even in the unaffected LH.CMRO2 estimates showed that metabolism wasmarkedly reduced in both mismatch and coretissues.

Magnetic resonance imaging-based CMRO2 modelremains to be validated. Davis’ CMRO2 model hasnot been previously used in disease states. Althoughthe stimulus-evoked CMRO2 changes and the Mvalues in the ischemic regions herein were reason-able and consistent, the Davis’ CMRO2 model maynot be valid under physiologically perturbed states.Direct validation of this MRI-based CMRO2 techni-que under normal and disease states is difficultbecause of a lack of gold standard for in vivo CMRO2

measurement. Cross correlation with microPET-CMRO2 techniques would provide further supportof this valuable MRI-based CMRO2 technique.Nonetheless, estimates of CMRO2 changes in strokeusing this model, particularly in the ‘perfusion–diffusion mismatch’ where the coupling couldremain largely intact, yielded interesting and valu-able information regarding tissue metabolism andthe underlying vascular coupling.

Conclusion

The MRI protocol and data analysis approachshowed herein offer a useful means to characterizeacute and chronic ischemic brain injury withrespect to anatomical, physiologic and functionalstatus at reasonably high spatiotemporal resolution.Future studies will make use of this approach, alongwith immunohistological and behavioral assess-ment, to systemically study ischemic brain injurywith variable functional outcomes (i.e., by increas-ing the transient ischemic duration) and wherereperfusion will be performed while the animalsare in the magnet to obtain a preocclusion dataset for robust pixel-by-pixel analysis. This approachhas the potential to be translated into clinicalsettings.

References

Astrup J, Symon L, Siesjo BK (1981) Thresholds incerebral ischemia: the ischemic penumbra. Stroke12:723–5

Austin VC, Blamire AM, Allers KA, Sharp T, Styles P,Matthews PM, Sibson NR (2005) Confounding effectsof anesthesia on functional activation in rodent brain:a study of halothane and a-chloralose anesthesia.NeuroImage 24:92–100

Binkofski F, Seitz RJ (2004) Modulation of the BOLD-response in early recovery from sensorimotor stroke.Neurology 63:1223–9

Boxerman JL, Hamberg LM, Rosen BR, Weisskoff RM(1995) MR contrast due to intravascular magneticsusceptibility perturbations. Magn Reson Med 34:555–66


1277


Calamante F, Williams SR, van Bruggen N, Kwong KK,Turner R (1996) A model for quantification of perfusionin pulsed labeling techniques. NMR Biomed 9:79–83

Corfield DR, Murphy K, Josephs O, Adams L, Turner R(2001) Does hypercapnia-induced cerebral vasodilationmodulate the hemodynamic response to neural activa-tion? Neuroimage 13:1207–11

Davis TL, Kwong KK, Weisskoff RM, Rosen BR (1998)Calibrated functional MRI: mapping the dynamics ofoxidative metabolism. Proc Natl Acad Sci USA95:1834–9

Dijkhuizen RM, Ren J, Mandeville JB, Wu O, Ozdag FM,Moskowitz MA, Rosen BR, Finklestein SP (2001)Functional magnetic resonance imaging of reorganiza-tion in rat brain after stroke. Proc Natl Acad Sci USA98:12766–71

Dijkhuizen RM, Singhal AB, Mandeville JB, Wu O,Halpern EF, Finklestein SP, Rosen BR, Lo EH (2003)Correlation between brain reorganization, ischemicdamage, and neurologic status after transient focalcerebral ischemia in rats: a functional magneticresonance imaging study. J Neurosci 23:510–7

Duong TQ, Iadacola C, Kim S-G (2001) Effect of hyperoxia,hypercapnia and hypoxia on cerebral interstitial oxy-gen tension and cerebral blood flow in the rat brain: an19F/1H study. Magn Reson Med 45:61–70

Duong TQ, Silva AC, Lee S-P, Kim S-G (2000) FunctionalMRI of calcium-dependent synaptic activity: crosscorrelation with CBF and BOLD measurements. MagnReson Med 43:383–92

Elsersy H, Sheng H, Lynch JR, Moldovan M, Pearlstein RD,Warner DS (2004) Effects of isoflurane versus fentanyl-nitrous oxide anesthesia on long-term outcome fromsevere forebrain ischemia in the rat. Anesthesiology100:1160–6

Grubb RL, Raichle ME, Eichling JO, Ter-Pogossian MM(1974) The effects of changes in PaCO2 on cerebralblood volume, blood flow, and vascular mean transittime. Stroke 5:630–9

Guadagno JV, Calautti C, Baron JC (2003) Progress inimaging stroke: emerging clinical applications. Br MedBull 65:145–57

Hoge RD, Atkinson J, Gill B, Crelier GR, Marrett S, Pike GB(1999) Investigation of BOLD signal dependence oncerebral blood flow and oxygen consumption: thedeoxyhemoglobin dilution model. Magn Reson Med42:849–63

Hossmann K-A (1994) Viability thresholds and thepenumbra of focal ischemia. Ann Neurol 36:557–65

Kety SS, Schmidt CF (1948) The effects of altered arterialtensions of carbon dioxide and oxygen on cerebralblood flow and cerebral oxygen consumption of normalyoung men. J Clin Invest 27:484–91

Kim S-G, Ugurbil K (1997) Comparison of blood oxygena-tion and cerebral blood flow effects in fMRI: estimationof relative oxygen consumption change. Magn ResonMed 38:59–65

Lees P (1972) Pharmacology and toxicology of alphachloralose: a review. Vet Rec 91:330–3

Li F, Han SS, Tatlisumak T, Liu KF, Garcia JH, Sotak CH,Fisher M (1999) Reversal of acute apparent diffusioncoefficient abnormalities and delayed neuronal deathfollowing transient focal cerebral ischemia in rats. AnnNeurol 46:333–42

Li F, Silva MD, Sotak CH, Fisher M (2000) Temporalevolution of ischemic injury evaluated with diffusion-,perfusion-, and T2-weighted MRI. Neurology 54:689–96

Liu ZM, Schmidt KF, Sicard KM, Duong TQ (2004a)Imaging oxygen consumption in forepaw somatosen-sory stimulation in rats under isoflurane anesthesia.Magn Reson Med 52:277–85

Liu ZM, Shen Q, Sicard KM, Febo M, Ferris CF, Stein EA,Duong TQ (2004b) Imaging cocaine-induced changes inBOLD, CBF and oxygen consumption. Proc Int SocMagn Reson Med 1:278

Mandeville JB, Marota JJ, Ayata C, Moskowitz MA,Weisskoff RM, Rosen BR (1999) MRI measurement ofthe temporal evolution of relative CMRO2 during ratforepaw stimulation. Magn Reson Med 42:944–51

Mandeville JB, Marota JJ, Kosofsky BE, Keltner JR,Weissleder R, Rosen BR, Weisskoff RM (1998) Dynamicfunctional imaging of relative cerebral blood volumeduring rat forepaw stimulation. Magn Reson Med39:615–24

Meng X, Fisher M, Shen Q, Sotak CH, Duong TQ (2004)Characterizing the diffusion/perfusion mismatch inexperimental focal cerebral ischemia. Ann Neurol55:207–12

Mortier E, Struys M, Herregods L (2000) Therapeutic comaor neuroprotection by anaesthetics. Acta Neurol Belg100:225–8

Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L,Shimizu H, Kucharczyk J, Wendland MF, WeinsteinPR (1990) Early detection of regional cerebral ischemiain cats: comparison of diffusion- and T2-weighted MRIand spectroscopy. Magn Reson Med 14:330–46

Novack P, Shenkin H, Bortin L, Goluboff B, Soffe AM(1953) The effects of carbon dioxide inhalation uponthe cerebral blood flow and cerebral oxygen consump-tion in vascular disease. J Clin Invest 32:696–702

Ogawa S, Lee T-M, Kay AR, Tank DW (1990) Brainmagnetic resonance imaging with contrast dependenton blood oxygenation. Proc Natl Acad Sci USA 87:9868–72

Ogawa S, Menon RS, Tank DW, Kim S-G, Merkle H,Ellermann JM, Ugurbil K (1993) Functional brainmapping by blood oxygenation level-dependent con-trast magnetic resonance imaging. Biophys J 64:800–12

Parkes LM, Tofts PS (2002) Improved accuracy of humancerebral blood perfusion measurements using arterialspin labeling: accounting for capillary water perme-ability. Magn Reson Med 48:27–41

Pineiro R, Pendlebury S, Johansen-Berg H, Matthews PM(2002) Altered hemodynamic responses in patientsafter subcortical stroke measured by functional MRI.Stroke 33:103–9

Roussel SA, van Bruggen N, King MD, Gadian DG (1995)Identification of collaterally perfused areas followingfocal cerebral ischemia in the rat by comparison ofgradient echo and diffusion-weighted MRI. J CerebBlood Flow Metab 15:578–86

Roy CS, Sherrington CS (1890) On the regulation of bloodsupply of the brain. J Physiol 1:85–108

Schmidt-Kastner R, Grosse Ophoff B, Hossman KA (1986)Delayed recovery of CO2 reactivity after one hour’scomplete ischaemia of cat brain. J Neurol 233:367–9

Shen Q, Meng X, Fisher M, Sotak CH, Duong TQ (2003)Pixel-by-pixel spatiotemporal progression of focalischemia derived using quantitative perfusion anddiffusion imaging. J Cereb Blood Flow Metab 23:1479–88

Shen Q, Ren H, Bouley J, Fisher M, Duong TQ (2004)Dynamic tracking of acute ischemic tissue fatesusing improved unsupervised ISODATA analysis of


1278


high-resolution quantitative perfusion and diffusiondata. J Cereb Blood Flow Metab 24:887–97

Sicard KM, Duong TQ (2005) Effects of hypoxia, hyperoxiaand hypercapnia on baseline and stimulus-evokedBOLD, CBF and CMRO2 in spontaneously breathinganimals. NeuroImage 25:850–8

Silva A, Williams D, Koretsky A (1997) Evidence for theexchange of arterial spin-labeled water with tissuewater in rat brain from diffusion-sensitized measure-ments of perfusion. Magn Reson Med 38:232–7

Silva AC, Lee S-P, Iadecola C, Kim S-G (2000) Earlytemporal characteristics of CBF and deoxyhemoglobinchanges during somatosensory stimulation. J CerebBlood Flow Metab 20:201–6

Silva AC, Lee S-P, Yang C, Iadecola C, Kim S-G (1999)Simultaneous blood oxygenation level-dependent andcerebral blood flow functional magnetic resonanceimaging during forepaw stimulation in the rat. J CerebBlood Flow Metab 19:871–9

Silva AC, Zhang W, Williams DS, Koretsky AP (1995)Multi-slice MRI of rat brain perfusion during amphe-

tamine stimulation using arterial spin labeling. MagnReson Med 33:209–14

Stejskal EO, Tanner JE (1965) Spin diffusion measure-ments: spin echoes in the presence of a time-dependentfield gradient. J Chem Phys 42:288–92

Ueki M, Linn F, Hossmann K-A (1988) Functionalactivation of cerebral blood flow and metabolism beforeand after global ischemia of rat brain. J Cereb BloodFlow Metab 8:486–94

van der Toorn A, Dijkhuizen RM, Tulleken CA, Nicolay K(1995) T1 and T2 relaxation times of the major 1H-containing metabolites in rat brain after focal ischemia.NMR Biomed 8:245–52

Weiss HR, Buchweitz E, Sinha AK (1983) Effectof hypoxic-hypercapnia on cerebral regionaloxygen consumption and supply. Circ Res 51:494–503

Zhou J, Wilson DA, Ulatowski JA, Traystman RJ, van ZijlPC (2001) Two-compartment exchange model forperfusion quantification using arterial spin tagging.J Cereb Blood Flow Metab 21:440–55


1279


Documents

Functional, perfusion and diffusion MRI of acute focal ... · Combined functional, perfusion and diffusion magnetic resonance imaging (MRI) with a temporal resolution of 30mins was