Hemodynamic and metabolic responses to activation, deactivation and epileptic discharges

www.elsevier.com/locate/ynimg

NeuroImage 28 (2005) 205 – 215

Hemodynamic and metabolic responses to activation, deactivation

and epileptic discharges

Bojana Stefanovic,* Jan M. Warnking, Eliane Kobayashi, Andrew P. Bagshaw, Colin Hawco,

Francois Dubeau, Jean Gotman, and G. Bruce Pike

Montreal Neurological Institute, 3801 University Street, Montreal, QC, Canada H3A 2B4

Received 2 February 2005; revised 24 March 2005; accepted 19 May 2005

Available online 5 July 2005

To investigate the coupling between the hemodynamic and metabolic

changes following functional brain activation as well as interictal

epileptiform discharges (IEDs), blood oxygenation level dependent

(BOLD), perfusion and oxygen consumption responses to a unilateral

distal motor task and interictal epileptiform discharges (IEDs) were

examined via continuous EEG-fMRI. Seven epilepsy patients per-

formed a periodic (1 Hz) right-hand pinch grip using ¨8% of their

maximum voluntary contraction, a paradigm previously shown to

produce contralateral M1 neuronal excitation and ipsilateral M1

neuronal inhibition. A multi-slice interleaved pulsed arterial spin

labeling and T2*-weighted gradient echo sequence was employed to

quantify cerebral blood flow (CBF) and BOLD changes. EEG was

recorded throughout the imaging session and reviewed to identify the

IEDs. During the motor task, BOLD, CBF and cerebral metabolic rate

of oxygen consumption (CMRO2) signals increased in the contra- and

decreased in the ipsilateral primary motor cortex. The relative changes

in CMRO2and CBF were linearly related, with a slope of 0.46 T 0.05.

The ratio of contra- to ipsilateral CBF changes was smaller in the

present group of epilepsy patients than in the healthy subjects

examined previously. IEDs produced both increases and decreases in

BOLD and CBF signals. In the two case studies for which the

estimation criteria were met, the coupling ratio between IED-induced

CMRO2and CBF changes was estimated at 0.48 T 0.17. These findings

provide evidence for a preserved coupling between hemodynamic and

metabolic changes in response to both functional activation and, for the

two case studies available, in response to interictal epileptiform activity.

D 2005 Elsevier Inc. All rights reserved.

Keywords: EEG-fMRI; Epilepsy; Negative BOLD; Perfusion; Oxygen

consumption

Introduction

Although the relationship between ictal and interictal epileptic

activity is not entirely understood (Badier and Chauvel, 1985,

1995; Alarcon et al., 1994; de Curtis and Avanzini, 2001; Avoli,

1053-8119/$ - see front matter D 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.neuroimage.2005.05.038

* Corresponding author. Fax: +1 514 398 2975.

E-mail address: [email protected] (B. Stefanovic).

Available online on ScienceDirect (www.sciencedirect.com).

2001; Janszky et al., 2001), interictal epileptiform discharges

(IEDs) represent a very specific marker of epilepsy, the delineation

of the irritative zone (Rosenow and Luders, 2001) being of

particular interest for presurgical evaluations of epileptic patients

(Penfield and Jasper, 1954; Kanner et al., 1995; McKhann et al.,

2000). Interictal activity has traditionally been studied with

electroencephalography (EEG), IEDs producing pronounced and

stereotyped electroencephalographic trace deviations. Although

ictal activity is generally associated with increased metabolism and

perfusion (Duncan, 1997), no consistent changes in cerebral

metabolic rate of glucose consumption (CMRGlc) or cerebral blood

flow (CBF) in response to interictal epileptiform activity have been

demonstrated (Theodore et al., 1985; Ochs et al., 1987). This has

often been ascribed to the poor sensitivity of the methods

employed, e.g., low statistical power and poor temporal resolution

of positron emission tomography (PET) studies, the latter leading

to amalgamation of different states within each measurement

(Duncan, 1997; Sperling and Skolnick, 1995).

In the last decade, functional magnetic resonance imaging

(fMRI) has been deployed in conjunction with EEG (Ives et al.,

1993; Huang-Hellinger et al., 1995; Warach et al., 1996) to

improve the EEG-based localization of the irritative zone and

enable investigation of the hemodynamic and metabolic correlates

of IEDs with high spatial and temporal resolution (Seeck et al.,

1998; Krakow et al., 1999; Patel et al., 1999; Lazeyras et al., 2000;

Jager et al., 2002; Al-Asmi et al., 2003). However, full use of the

information afforded from fMRI BOLD data in the combined

EEG-fMRI investigations is predicated on the understanding of the

physiological changes determining the BOLD response, which are

incompletely understood even in normal functional brain activa-

tion. Detailed investigation of the BOLD response to IEDs has

only recently begun (Lemieux et al., 2001; Benar et al., 2002;

Bagshaw et al., 2004; Aghakhani et al., 2004). Regional negative

BOLD responses to IEDs have been observed (Salek-Haddadi

et al., 2003b; Archer et al., 2003a,b; Bagshaw et al., 2004;

Aghakhani et al., 2004), but their origins are presently unknown.

In contrast to focal epilepsies where an uncoupling between

CBF and CMRGlc has been suggested (Gaillard et al., 1995; Fink

et al., 1996; Breier et al., 1997; Bruehl et al., 1998) (and disputed

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215206

(Franck et al., 1989; Kuhl et al., 1980)), the diffuse neuro-

physiological abnormalities assumed to exist in patients with

idiopathic generalized epilepsy (IGE) are commonly believed not

to influence the nature of the coupling between metabolic and

hemodynamic changes, and hence the interpretation of BOLD

fMRI response. While the neurovascular coupling in IGE has

largely been unexplored, it may, in principle, still be compromised

(as it is in partial epilepsies) even in the interictal state as a result of

pharmacological interventions (Theodore et al., 1989; Leiderman

et al., 1991), the cumulative effect of seizures, an IED-instigated

rise in extracellular potassium (Jensen and Yaari, 1997), pH

changes in the course of after-potentials hyperpolarization (de

Curtis et al., 1998) and compromised functioning of astrocytes

(Grisar et al., 1999), potential mediators of neurovascular coupling

(Salek-Haddadi et al., 2003a).

In the present study, we set out to investigate the coupling

between perfusion and oxygen consumption changes in a group of

epilepsy patients, in response to both a motor task and IEDs.

Although no direct measurements of this coupling have thus far

been made, it is critical for the interpretation of the BOLD fMRI

studies in epilepsy. The selected motor task, involving a low-force

phasic pinch grip, is known to produce neuronal deactivation (and

correspondingly, a negative BOLD response) in the ipsilateral

primary motor cortex in addition to neuronal activation (producing

a BOLD signal increase) of the contralateral primary motor cortex

(Boroojerdi et al., 1996; Netz et al., 1995; Gerloff et al., 1998;

Liepert et al., 2001; Ferbert et al., 1992; Hamzei et al., 2002). In a

recent study of healthy young adults performing this task, we

found the same coupling between changes in oxygen consumption

and flow in response to both neuronal activation and deactivation

(Stefanovic et al., 2004). In view of the present understanding of

IGE, we hypothesized that these patients would exhibit normal

coupling between hemodynamic and metabolic responses follow-

ing neuronal activation or deactivation, with a consistent relation-

ship between CMRO2and CBF changes underlying both positive

and negative steady-state BOLD signal changes. Furthermore, the

relatively small perturbations in neuronal state elicited by IEDs

were expected to fall within the normal operating range of the

neuronal circuitry: the spike-induced flow and oxygen consump-

tion changes were thus anticipated to follow those observed during

physiologic stimulation.

Here, we report on the relationship between blood flow and

oxygen consumption in regions of sustained BOLD and CBF

signal increases, in the contralateral primary motor cortex (M1),

and their decreases, in the ipsilateral M1, in response to the motor

task previously employed in a study of healthy volunteers.

Furthermore, the hemodynamic and metabolic responses induced

by interictal epileptiform activity were quantified. The present

findings elucidate the nature of the metabolic and vascular

responses to normal neuronal activation, deactivation and IEDs

in the brain of patients with epilepsy.

Methods

Overall, the BOLD and CBF responses were simultaneously

measured in a group of epilepsy patients exhibiting generalized

IEDs who performed a motor task known to produce neuronal

activation in the contra- and deactivation in the ipsilateral primary

motor cortex. The electroencephalogram was recorded simulta-

neously with the fMRI to identify the occurrences of the IEDs and

thus allow the quantification of BOLD and CBF responses to IEDs.

The concomitant changes in oxygen consumption were estimated

via the deoxyhemoglobin dilution model (Davis et al., 1998; Hoge

et al., 1999) in conjunction with BOLD and CBF data obtained

during the administration of hypercapnia.

Motor task

The motor task employed was identical to that described in an

earlier study of 8 healthy adults (Stefanovic et al., 2004). Briefly,

the maximum voluntary contraction of the right-handed pinch grip

was measured for each patient. The patients were then trained to

perform the pinch grip at a frequency of 1 Hz, on each grip

pressing a water-filled ball with the thumb and the index finger of

the right hand. The recording and analysis of the exerted pressure

was performed in real time and an auditory feedback provided to

the patient: a low frequency tone indicated that the force applied

was in the desired range, namely within 20% of the target level. A

high frequency tone accompanied too strong a force, and no tone

was played out when insufficient force was exerted. To minimize

habituation, the target level was randomized on each pinch grip and

varied between 6 and 10% of the patient’s maximum voluntary

contraction.

Hypercapnic modulation

Mild hypercapnia was induced through administration of

mixtures of carbon dioxide and air through a non-rebreathing face

mask (Hudson RCI, Model 1069, Temecula, CA). At baseline, the

subjects were inhaling medical air (21% O2, balance N2), supplied

at 16 L/min. During hyper-capnic perturbations, a premixed

preparation of 10% CO2, 21% O2 and balance N2 (BOC Canada

Ltd., Montreal, Quebec, Canada) was combined with medical air in

a Y-connector. Three levels of hypercapnia were administered, with

CO2 concentration of 4, 6, and 8%, respectively. At each level, the

gas flow rates were adjusted to maintain a total flow rate of 16 L/

min. End-tidal CO2 was measured via a nasal cannula with

monitoring aspirator (Normocap 200, Datex Inc., Plymouth, ME)

and increased an average 17 T 4 mm Hg (or 59 T 21%) during

inhalation of the highest CO2 concentration mixture. Subjects were

asked to breathe at a constant rate, and their respiratory rate was

monitored via the capnometer.

Patients

Seven patients (6 females and 1 male; average age 40 T 6 years)

participated in the study. They were selected from a database of 91

epileptic patients (of which 25 had IGE) who had undergone

combined EEG-fMRI monitoring at the Montreal Neurological

Institute. The patients were selected based on a strong IED-induced

negative BOLD response (at least one cluster of �20 voxels with

peak t value ��5) and overall cooperativity in the prior (no more

than 30 months earlier) scanning session, as well as willingness to

participate. The exclusion criteria comprised a history of asthma or

past neurosurgical interventions. Six of the seven patients were

clinically diagnosed with IGE and the remaining one with

parietooccipital epilepsy. The top portion of Table 1 summarizes

the clinical characteristics of each patient. Informed consent was

obtained from each subject prior to the scanning session, the

experimental protocol having been approved by the Research

Ethics Board of the Montreal Neurological Institute.

Table 1

Summary of clinical characteristics, electroencephalographic findings, and BOLD/CBF regions of activation/deactivation in each subject

Subject Id 1 2 3 4 5 6 7

Age (at onset) [years] 43 (8) 34 (7) 67 (19) 24 (7) 51 (4) 26 (12) 38 (14)

Sex F F M F F F F

Epilepsy syndrome POE JAE JME JAE JAE JAE JME

Medications [mg/day] CBZ 800 LEV 2000 LMT 400 LMT 500 LMT 200 VA 1000 CBZ 200

LMT 400 TPM 50 VA 750 VA 375 VA 500

VA 1750

Anatomical MRI Bilateral PO

polymicrogyria

Normal Enlarged ventricles Normal Normal Normal Normal

SW frequency [Hz] polyspikes 2–3 4 3–4 3 N/A 4

No. of events (bursts) 404 (N/A) 82 (5) 22 (1) 39 (8) 33 (14) 0 (0) 40 (7)

Burst duration [s] <1 4.0 T 1.8 2 T 0 2.7 T 0.8 2.5 T 0.8 N/A 2.2 T 0.3

BOLD activations b m PO r PO/b m P/b TPO/b TPO b C l C b m PO; b P None b F; b O

BOLD deactivations b P; b Pc none/b PO/b m O/b m PO r m P; r Mc; b P; l PO r F; l C None None b Pc; b PO

CBF activations b FP b P; b PO F; b FP; b P A r F A None b PO

CBF deactivations b PO b F; l PO; b m PO; b m PO A b F; b P l F None b F

SW: spike and wave; POE: parietooccipital epilepsy; JAE: juvenile absence epilepsy; JME: juvenile myoclonic epilepsy; CBZ: carbamazepine; LMT:

lamotrigine; LEV: levetiracetam; TPM: topamax; VA: valproate; b: bilateral; r: right; l: left; m: mesial; P: parietal; O: occipital; F: frontal; T: temporal; C:

central; Pc: posterior cingulate; Mc: mid-cingulate; A: scattered throughout the brain. Note that the number of events comprises both isolated events and bursts.

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215 207

Experiment

The scanning protocol consisted of a high-resolution 3 D RF-

spoiled T1-weighted gradient echo (1 � 1 � 2 mm3) sequence for

anatomical reference, followed by a multi-slice interleaved pulsed

arterial spin labeling (PASL) and T2*-weighted gradient echo

sequence for CBF and BOLD signal measurements. The high-

resolution gradient echo sequence employed a TR of 22 ms, a TE

of 10 ms and non-selective 30- RF-spoiled excitation. The CBF

and BOLD acquisitions covered 8 slices (inter-slice gap of 1 mm),

positioned to include the primary motor cortices as well as the

regions showing most prominent BOLD activations and deactiva-

tions based on the results of the prior EEG-fMRI experiment. The

CBF data were acquired using a QUIPSS II sequence (Wong et al.,

1997) with two presaturation BASSI pulses (Warnking and Pike,

2004) in the imaging region followed by an adiabatic BASSI

inversion pulse in the labeling region (thickness of 100 mm, gap of

5 mm), with a QUIPSS II delay (TI1) of 700 ms and a post-label

delay (TI2) of 1300 ms. An EPI readout (2232 Hz/pixel) was

employed, with an echo time of 50 ms for BOLD and 28 ms for

CBF (with 4 � 4 � 5 mm3 resolution). To increase the signal-to-

noise ratio, the resolution was decreased to 5 � 5 � 5 mm3 for the

latter 5 subjects, resulting in a 22-ms TE for CBF. The repetition

time was 2.5 s for the first 2 subjects (subjects 2 and 7), and 1.8 s

for the remaining 5 subjects. EEG data were continuously recorded

via 21 MR-compatible Ag/AgCl electrodes, placed according to

the International 10–20 system, with signal sampled at 5 kHz and

amplified using a BrainAmp (Brain Products, Munich, Germany)

amplifier.

The functional paradigm involved 24 repetitions of ¨20 s/60 s/

40 s off/on/off blocks (only 12 repetitions were done in subjects 2

and 7), rest alternating with the low force, phasic, right-handed

pinch grip, the beginning of each block being indicated by

auditory cues. Following the functional scan, medical air alter-

nating with graded hypercapnia was administered in three 1 min/3

min/2 min blocks. The total scanning time was approximately 75

min. Patients were immobilized using a vacuum bag (S&S x-ray

products, Brooklyn, NY). The RF body coil was used for

transmission and a quadrature head coil for signal reception. All

the examinations were performed on a Siemens 1.5 T Magnetom

Sonata system.

Data analysis

EEG data analysis

The EEG was filtered offline to remove the artifact generated

by the MR scanner (Vision Analyser software, BrainProducts,

Munich, Germany). EEG was reviewed by an experienced

electroencephalographer (EK), who identified the IEDs according

to their spatial distribution, morphology, timing and duration. The

durations of the events marked were not different from habitual

IEDs. The type of IED was determined based on the spatial

distribution and morphology observed on the scalp EEG, resulting

in 10 data sets from the seven patients (subject 6 had no IEDs and

subject 2 had 4 event types).

fMRI data statistical analysis

The fMRI data were motion corrected using AFNI’s 3dvolreg

software (Cox, 1996). The data were spatially smoothed using a

three dimensional Gaussian filter with full width half maximum of

6 mm. Drift was removed by subtracting from each voxel’s time

course the low-frequency components of its discrete cosine

transform, with a cutoff frequency of one half of the stimulation

paradigm frequency. The generalized linear model (Worsley et al.,

2002) was used to identify areas of statistically significant task

correlation as well as those of statistically significant IEDs

correlation at the omnibus significance level of 0.05 in BOLD

and CBF data, respectively. The motor task and hypercapnia

induced BOLD and CBF response amplitudes from the general

linear model were normalized by the baseline signal values and

averaged over the voxels of each ROI. To allow for establishment

of a physiological steady-state, the hypercapnic data acquired

within half a minute following a change in the concentration of the

inspired CO2 was excluded from the analysis.

The ratios of contra- to ipsilateral BOLD and CBF responses to

the motor task were compared to the corresponding values in

healthy controls (Stefanovic et al., 2004). Notably, the normal-

ization of contra- by ipsilateral values was performed in each

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215208

subject prior to the comparison in view of the potential confounding

effects of the antiepileptic medications on the global resting flow

(Gaillard et al., 1996) and the known large inter-subject variability

in CBF responses to motor activation in healthy subjects.

With respect to the IED events, each data set was analyzed with

four impulse response functions, modeled as monophasic, single

gamma functions, peaking 3, 5, 7 and 9 s following the event to

allow for some variation in the latency of the response while

retaining information about its expected shape (Buckner, 1998;

Bagshaw et al., 2004). The full width at half maximum of the

gamma functions was 5.2 s; the duration of the burst was input

when modeling the response. Composite statistical maps were

created by taking the maximum value from the four analyses at

each voxel, as described by Bagshaw et al. (2005).

IED response estimation

The regions of interest for IED response estimation were chosen

as the clusters in BOLD/CBF t value maps showing statistically

significant correlation with the IED events and having at least 5

and no more than 20 voxels (i.e., with a volume, V, 0.625 cc � V �2.5 cc). The IED responses were then estimated by fitting the fMRI

signal within each region of interest (ROI) using a Fourier basis set

(Josephs et al., 1997; Kang et al., 2003). Only data from those

BOLD/CBF ROI pairs whose centers of mass were within 5 mm of

each other were considered for CMRO2estimation, described

below. Within each subject, the peak BOLD/CBF changes from all

regions of interest satisfying this criterion were averaged before the

CMRO2estimation was done.

CMRO2estimation

The hypercapnic data were averaged across all subjects, at each

level of hypercapnia and a common maximum achievable BOLD

signal change (M) was estimated by linear fitting of the trans-

formed and averaged CBF data vs. averaged BOLD data to the

deoxyhemoglobin dilution model (Davis et al., 1998; Hoge et al.,

1999):

DBOLD

BOLD0

¼ M 1� CBF

CBF0

�� a�h��: ð1Þ

We thus assumed no effect of the mild hypercapnia elicited in

this experiment on the rate of oxygen consumption. The a and h

Fig. 1. Regions of interest based on t value maps thresholding for BOLD (a) and C

are overlaid on the average of all subjects anatomical scans in the Talairach spac

were set to 0.38 and 1.5, respectively (Grubb et al., 1974;

Boxerman et al., 1995). The individual motor task-induced CMRO2

changes were next calculated using the M (and its associated

standard error) from Eq. (1) in combination with the measured

BOLD and CBF data during the functional run, as follows (Davis

et al., 1998; Hoge et al., 1999):

CMRO2

CMRO2j0

¼ 1�DBOLDBOLD0

��M

1A

0@

1h

CBF

CBF0

�� 1� ah

: ð2Þ

Therefore, the errors in the M estimate from the linear fitting of

the transformed and averaged CBF hypercapnia data to averaged

BOLD hypercapnia data were propagated into the errors on the

calculated activation-induced CMRO2changes. Finally, a single

straight line was fit to the noisy CMRO2, noisy CBF data pairs from

both contra- and ipsilateral ROIs of all subjects to obtain an

optimal estimate of the CMRO2/CBF coupling ratio for the motor

task. The same analysis was performed for the peak BOLD and

CBF changes from IED-elicited responses to estimate CMRO2

change as well as the CMRO2/CBF coupling ratio following IEDs.

The quality of each fit was assessed by v2 analysis, with the v2

probability reported as q (Press et al., 1992).

Results

Motor task-induced responses

Task-induced increases in BOLD signal were observed,

contralaterally, in the primary sensorimotor cortex (SM1), premo-

tor cortex (PMC), supplementary motor area (SMA), as well as a

part of the posterior parietal association cortex (PPC) flanking the

postcentral sulcus. Ipsilaterally, BOLD signal increased in the

secondary areas (namely, PMC, SMA and PPC), but decreased in

the primary sensorimotor cortex. Fig. 1 shows a slice of BOLD and

CBF ROIs, summed over all subjects after registration (Collins et

al., 1994) with the Montreal Neurological Institute template brain

(Evans et al., 1993). The (x, y, z) coordinates (in mm) of the center

of mass of the t value based primary motor cortex ROIs

transformed into the Talairach space and summed over all subjects

were (�35, �24, 54) for contralateral BOLD, (36, �23, 55) for

BF (b), transformed into the Talairach space and summed over all subjects,

e. The contralateral ROIs are displayed in red and the ipsilateral in green.

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215 209

ipsilateral BOLD, (�37, �22, 53) for contralateral CBF, and (35,

�24, 54) for ipsilateral CBF.

A typical set of BOLD signal and CBF time courses, in both

contra- and ipsilateral M1-ROIs of a subject, is shown in Fig. 2.

Fig. 3 displays all measured BOLD and CBF data pairs, for

hypercapnic perturbation and motor task, as well as the calculated

iso-CMRO2contours. In 5 out of 7 subjects, the magnitude of CBF

and BOLD signal changes were larger in the contra- than in the

ipsilateral ROI.

The maximum achievable BOLD signal increase (M), obtained

by linear fitting of the average hypercapnia data across all subjects,

was 0.046 T 0.013, corresponding to aDR2* of�0.9 T 0.2 s1. The v2

analysis indicated a good fit ( q = 0.39) (Press et al., 1992). The

calculated CMRO2and the corresponding measured CBF changes,

for each subject, are displayed in Fig. 4. The slope of the straight line

fit to these data yielded a CMRO2/CBF coupling ratio of 0.46 T 0.05

(with q of 0.92 indicating an excellent fit (Press et al., 1992)).

Comparison with controls

The data of Fig. 3 has been replotted in Fig. 5, in conjunction

with the corresponding data from our previous study, which

Fig. 2. Time courses of contralateral (positive) BOLD (a) and CBF (b), as well as ip

standard errors are shown as dashed lines. All time course data have been low pas

24 sessions.

employed the same motor paradigm in a group of healthy subjects

(Stefanovic et al., 2004). Notably, the ratio of contra- to ipsilateral

CBF responses is significantly (P ¨ 0.017) smaller in epileptic

patients (2.2 T 1.3) than in healthy subjects (4.2 T 2.3). Average

contra- and ipsilateral CBF responses across subjects are shown in

Fig. 6. A very similar level of deactivation-induced percent CBF

decrease is seen in the two groups, in contrast to a smaller

excitation-induced percent CBF increase in the epilepsy patients.

The same trend of lower contra-to ipsilateral BOLD responses in

epileptic patients relative to controls is observed (2.7 T 2.5 vs. 3.7 T2.1), though it does not reach significance (P ¨ 0.21).

IED-induced responses

Six out of seven patients exhibited epileptiform activity in the

course of the scanning session (cf. Table 1 for the summary of EEG

findings). In each of these subjects, the interictal epileptiform

discharges induced both increases and decreases in BOLD and

CBF signals, as summarized in Table 1. Only two of the six

subjects exhibited sufficiently co-localized (i.e., at most 5 mm

separation between their respective ROI centers of mass, as

described in Methods) statistically significant changes in both

silateral (negative) BOLD (c) and CBF (d) percent changes in subject 1. The

s filtered with a Hanning window (FWHM = 20 s) prior to averaging across

Fig. 3. The percent changes in BOLD and CBF signals in the ipsilateral

ROIs (green circles) and contralateral ROIs (red triangles) for each subject.

The average hypercapnia data (black squares) are displayed along with the

corresponding fit (indicated by crosses), representing the baseline iso-

CMRO2contour. The estimate of the maximum achievable BOLD signal

change was substituted into the equation [13] of the deoxyhemoglobin

dilution model (Hoge et al., 1999) to generate non-baseline iso-CMRO2

contours (shown as solid black curves), at 10% intervals. The shaded area

corresponds to the shaded region of Fig. 4.

Fig. 5. The percent changes in BOLD and CBF signals induced by the

motor task in the ipsilateral ROIs (green circles: epilepsy patients; blue

circles: healthy subjects) and contralateral ROIs (red triangles: epilepsy

patients; magenta triangles: healthy subjects).

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215210

BOLD and CBF. These included right parietal and right cuneus

regions in subject 1; and bilateral frontal, left occipital, bilateral

precentral, left precuneus and right cuneus regions in subject 7.

The two BOLD/CBF ROI pairs in subject 1 and 13 BOLD/CBF

ROI pairs in subject 7 met the above criteria. Sample t value maps

from subject 1 are shown in Fig. 7. As described in Methods,

BOLD and CBF data were averaged across the ROIs of each

subject prior to the calculation of CMRO2changes. The optimal

linear fit between the resulting CMRO2estimates and CBF data in

the ROIs of these two subjects is displayed in Fig. 8. The slope of

the straight line fit to these data yielded a CMRO2/CBF coupling

Fig. 4. The oxygen consumption changes corresponding to each subject’s

perfusion changes induced by the motor task in the ipsilateral ROIs (green

circles) and contralateral ROIs (red triangles). The optimal straight line fit

(q = 0.92) to these data is shown superimposed, yielding a coupling ratio of

0.46 T 0.05. The shaded region represents the standard error in the linear fit.

ratio of 0.48 T 0.17 (with q of 0.80 indicating a very good fit (Press

et al., 1992)).

Discussion

The present experiments provide, for the first time, a complete

set of BOLD, CBF and CMRO2measurements following functional

activation, deactivation and IEDs in epilepsy patients. They

demonstrate a preserved coupling between perfusion and oxygen

consumption changes in epilepsy patients. As was the case for the

healthy volunteers, the CMRO2/CBF relationship was consistent

between regions of positive and negative BOLD responses to a

motor task (with DCMRO2/DCBF of 0.46 T 0.05). For the regions

that showed statistically significant IED-induced changes in both

BOLD and CBF (thus allowing for CMRO2estimation), a similar

CMRO2/CBF coupling ratio, of 0.48 T 0.17, was estimated. Overall,

these findings are consistent with the general notion of epilepsy as

Fig. 6. The average, motor task induced, percent changes in CBF signal in

ipsi- and contralateral ROIs of healthy subjects (C) and epileptic patients (E).

Fig. 7. Sample BOLD (left) and CBF (right) t value maps in a subject (Subject Id. 1), overlaid on the corresponding anatomical slices. The regions of positive

responses are shown in the top row; the regions of negative responses, in the bottom row. The centers of mass for the overlapping regions are shown with a

cross hair.

Fig. 8. The oxygen consumption changes corresponding to across ROI

average IED-induced perfusion and BOLD changes in subjects 1 and 7. The

averages for ROIs showing IED-induced signal decreases are shown as

green circles; the averages for ROIs showing IED-induced signal increases,

as red triangles. The optimal straight line fit (q = 0.80) to these data is

shown superimposed, yielding a coupling ratio of 0.48 T 0.17. The shaded

region represents the standard error in the linear fit.

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215 211

a disorder of neuronal excitability, involving neuronal disinhibition

and hyperexcitability.

While neuronal hyperexcitability is thought to characterize

most epilepsy syndromes, the pathophysiology of these diseases is

still incompletely understood. The role of genetics in idiopathic

generalized epilepsy has long been suspected (Metrakos and

Metrakos, 1961) and a number of different IGE subsyndromes

have recently been associated with distinct mutations in GABAA

receptor sub-units (Macdonald et al., 2004; Gutierrez-Delicado and

Serratosa, 2004). Although IGE patients have normal structural

MRI, regional decreases in their N-acetyl aspartate levels have

been reported (Savic et al., 2004), suggesting a heterogeneous,

diffuse neuronal abnormality. The average resting metabolism and

flow in IGE patients are largely unremarkable (Duncan, 1997;

Theodore et al., 1985; Ochs et al., 1987; Kapucu et al., 2003;

Devous et al., 1990), in sharp contrast to the hypometabolism and

hypoperfusion frequently observed in the area of the epileptogenic

focus, its immediate surround or elsewhere in the brain of patients

with localization-related epilepsies (Kuhl et al., 1980; Engel et al.,

1982; Lee et al., 1986; Franck et al., 1986; Kim et al., 2001). While

initial seizures in serial seizure animal models were accompanied

by the expected increases in cerebral blood volume, arterial blood

pressure, cortical oxygen tension and cytochrome oxidase pressure,

one or more of these variables failed to rise in response to

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215212

subsequent seizures, testifying to a gradual breakdown of neuro-

vascular coupling in these patients (Kreisman et al., 1981, 1983).

Similarly, in a near-infrared spectroscopy study of pediatric

epileptic seizures, an early CBV increase gradually changed to a

CBV decrease in the course of the seizure in a patient with tonic

status epilepticus (Haginoya et al., 2002). In line with the present

findings, these data suggest a preserved interictal neurovascular

coupling that is progressively compromised in the course of either

sustained or highly repetitive ictal events.

Each patient in this study was taking one or more antiepileptic

drugs (AEDs), which were reported to reduce baseline CMRGlc and

CBF (Theodore et al., 1989; Leiderman et al., 1991; Spanaki et al.,

1999; Gaillard et al., 1996). A reason for these reductions may lie

in the decreased metabolic requirements following the enhance-

ment of cerebral inhibitory neurotransmission (Theodore, 1988). In

view of the effect of valproate on brain and CSF GABA levels

(Loscher, 1979, 1981), it is important to note that increased CSF

GABA (which is linearly related to brain GABA (Palfreyman et al.,

1983; Petroff et al., 1996)), following administration of the GABA

agonist muscimol, was found to affect both blood flow and glucose

consumption as to maintain a normal relationship between the two

(Kelly and McCulloch, 1983). While the medications might have

shifted the absolute global flow and metabolism in these subjects,

there is no evidence in support of their influence on the relationship

between metabolic and hemodynamic responses to changes in

neuronal activity. Furthermore, we do not expect any spatial

variation of their effects across homologous brain regions—

namely, primary motor cortices—of interest for our functional

paradigm.

While fMRI BOLD studies of physiological brain activation are

often done in surgical epileptic candidates to map the eloquent

cortex (Deblieck et al., 2003; Diehl et al., 2003; Huettel et al.,

2004; Szaflarski et al., 2004), there are no studies making

simultaneous measurements of activation-induced BOLD, CBF

and CMRO2changes in epilepsy patients, likely due to the

complexity and limited SNR of such measurements. The slope of

0.46 T 0.05 of the best line fit to both contra- and ipsilateral M1

CMRO2vs. CBF percent signal changes found here is in excellent

agreement with the value of 0.44 T 0.04 we reported in an earlier

study of healthy volunteers performing the same motor task. It is

also in reasonable agreement (given the paradigm differences and

the expected intersubject variability) with the ratios reported by our

and other groups for the contralateral primary motor cortex

activation in studies of BOLD signal increases, with the average

of 0.35 T 0.03 found in this lab (Atkinson et al., 2000) and 0.33 T0.06 reported by Kastrup (Kastrup et al., 2002).

The ratio of the changes in perfusion—between the contrala-

teral region of neuronal activation and homologous ipsilateral

region of neuronal deactivation—was decreased in the epileptic

patients compared to the corresponding value in healthy subjects

studied previously. This significant difference in the CBF ratios

resulted from a decrease in the activation-induced CBF increases,

with a preserved range of deactivation-induced CBF decreases. In

view of the diffuse cortical hyperexcitability presumed to exist in

these patients and the suggested dominant contribution of

presynaptic potentials to the total metabolic demands of neuronal

activity (Logothetis et al., 2001), it is tempting to speculate that the

relative metabolic cost of neuronal activation with respect to

neuronal deactivation may be diminished in these patients when

compared to healthy volunteers, though the neuronal excitability

(and hence the energetic cost of the activation) may well be

influenced by AEDs (Tassinari et al., 2003). The dissociation of the

effects of the underlying pathology from those of the medications

is, however, presently unavailable.

In contrast to the sustained after-depolarizations and multiple

spike discharges characteristic of the ictal state (Matsumoto and

Ajmone Marsan, 1964a), IEDs are associated with a paroxysmal

depolarization shift of the resting neuronal membrane potential,

bursts of action potentials and ensuing hyperpolarization and hence

inhibition (Matsumoto and Ajmone Marsan, 1964b). In view of

this pronounced difference in the electrophysiological signature of

the two states, the CBF and CMRO2changes induced by interictal

discharges are expected to be far less conspicuous than their ictal

counterparts (Prevett et al., 1995; Engel et al., 1985; Theodore et

al., 1985). Likely due to the limited sensitivity of the methods in

combination with the sparcity of the interictal discharges in most

patients, there are few reports of metabolic and hemodynamic

changes induced by IEDs in epilepsy patients. No effect of the

spike and wave activity on the CMRGlc was observed in a PET

study of a group of generalized epilepsy patients, though there was

a slight trend toward CMRGlc increases in IGE patients (Ochs et al.,

1987). In a PET study of a reflex epilepsy patient, a 34.6% increase

in CBF and 13% increase in CMRGlc were measured in a region

concordant with the site of maximal ictal EEG abnormality, as

determined by implanted electrodes (Bittar et al., 1999). In a group

of patients with photosensitive epilepsy, a significant blood flow

increase was measured in the hypothalamus during a photo-

paroxysmal response (da Silva et al., 1999). In the caudate nucleus,

the CBF increase instigated by the intermittent photic stimulation

was abolished during the photoparoxysmal response (da Silva et

al., 1999).

Both widespread positive and negative BOLD responses to

IEDs have been reported in a number of EEG-fMRI studies (Salek-

Haddadi et al., 2003b; Archer et al., 2003a,b; Bagshaw et al., 2004;

Aghakhani et al., 2004). In the present study, we also observed

regions of both BOLD and CBF increases and decreases. Since

BOLD signal has a complex dependence on a number of

physiological parameters (Davis et al., 1998; Hoge et al., 1999),

we also estimated the corresponding oxygen consumption changes,

to obtain a more direct marker of the underlying metabolic costs.

Due to a combination of factors—the limited PASL contrast-to-

noise ratio at 1.5 T; few, transient IED events (compared to e.g.,

numerous repetitions of the block motor paradigm employed when

estimating the DCMRO2following neuronal activation/deactiva-

tion); and the stringent requirement for the overlap of BOLD/CBF

ROIs (to ensure robust, colocalized BOLD and CBF measure-

ments), the set of data available for quantification of oxygen

consumption changes in response to IEDs was severely curtailed.

A total of only 15 ROIs (showing statistically significant signal

changes)—from 2 patients—were sufficiently overlapped to allow

the estimation of the corresponding oxygen consumption. In these

2 case studies, BOLD changes were invariably accompanied by

CBF changes of the same sign, with the estimated DCMRO2/DCBF

coupling ratio of 0.48 T 0.17, thus very close to the one observed

for the functional activation in these patients as well as the one

obtained in healthy volunteers (Stefanovic et al., 2004). Nonethe-

less, the paucity of data available for this estimation, in com-

bination with the large variability of both epileptic syndromes and

the nature and dosage of medications customarily prescribed in its

treatment preclude any general conclusions about the hemody-

namic and metabolic responses to IEDs to be made from the

present results.

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215 213

While these data suggest that IED-induced negative BOLD

responses may arise from the larger flow relative to oxygen

consumption decreases, as observed for motor task-induced

negative BOLD responses in healthy volunteers (Stefanovic et

al., 2004), other explanations of negative BOLD phenomena are

still possible. This is particularly true of epileptogenic zones in

focal epilepsies and responses to ictal activity, where neurovascular

coupling may well be compromised, as suggested earlier. Never-

theless, it is tempting to apply Gloor’s account of the spike and

wave phenomenon (Gloor, 1978), thus hypothesizing that the

presently measured negative CBF and BOLD responses result from

the net deactivation of the region due to a locally predominant

cortical inhibition relative to excitation. This also allows for the

existence of regions showing no BOLD response, due to a balance

between changes in local excitation and inhibition, integrated over

the interval determined by the effective BOLD temporal resolution,

as proposed earlier (Archer et al., 2003a). Finally, the metabolic

costs of IEDs and hence the ensuing CMRO2response and,

indirectly, BOLD response are likely affected by the relative

contributions of changes in synchronicity vs. synaptic activity to

the generation of IEDs (Salek-Haddadi et al., 2003b).

Conclusion

We observed normal hemodynamic responses to hypercapnic

perturbation in a group of epilepsy patients with generalized IEDs.

A consistent linear relationship between oxygen consumption and

perfusion changes during motor task performance in regions of

sustained positive as well as negative BOLD response was found.

The slope of the linear fit to CMRO2vs. CBF changes from both

ipsi- and contralateral ROIs was 0.46 T 0.05, in close agreement

with the coupling ratio found in an earlier study of healthy

volunteers. On the other hand, a decreased ratio of the magnitude

of contra- to ipsilateral flow changes was observed in the patient

group. Interictal epileptiform discharges produced a similar

coupling, with DCMRO2/DCBF of 0.48 T 0.17. The current

findings suggest a preserved coupling between metabolic and

hemodynamic processes underlying BOLD increases and

decreases in epileptic patients, in response to both normal

functional activation and IEDs and provide no evidence for a

disturbance in the interictal cerebral vascular responses in this

disorder.

Acknowledgments

This work was supported by the Natural Sciences and

Engineering Research Council of Canada and the Canadian

Institutes of Health Research.

References

Aghakhani, Y., Bagshaw, A.P., Benar, C.G., Hawco, C., Andermann, F.,

Dubeau, F., Gotman, J., 2004. fMRI activation during spike and wave

discharges in idiopathic generalized epilepsy. Brain 127, 1127–1144.

Alarcon, G., Guy, C.N., Binnie, C.D., Walker, S.R., Elwes, R.D., Polkey,

C.E., 1994. Intracerebral propagation of interictal activity in partial

epilepsy: implications for source localisation. J. Neurol., Neurosurg.

Psychiatry 57 (4), 435–449.

Al-Asmi, A., Benar, C.G., Gross, D.W., Khani, Y.A., Andermann, F., Pike,

G.B., Dubeau, F., Gotman, J., 2003. fMRI activation in continuous and

spike-triggered EEG-fMRI studies of epileptic spikes. Epilepsia 44 (10),

1328–1339.

Archer, J.S., Abbott, D.F., Waites, A.B., Jackson, G.D., 2003a. fMRI

‘‘deactivation’’ of the posterior cingulate during generalized spike and

wave. NeuroImage 20 (4), 1915–1922.

Archer, J.S., Briellman, R.S., Abbott, D.F., Syngeniotis, A., Wellard, R.M.,

Jackson, G.D., 2003b. Benign epilepsy with centro-temporal spikes:

spike triggered fMRI shows somato-sensory cortex activity. Epilepsia

44 (2), 200–204.

Atkinson, J.D., Hoge, R.D., Gill, B., Sadikot, A.F., Pike, G.B., 2000.

BOLD, CBF, and CMRO2in the human primary motor cortex.

Proceedings of the Sixth International Conference on Functional

Mapping of the Human Brain, San Antonio, Elsevier, p. S803.

Avoli, M., 2001. Do interictal discharges promote or control seizures?

Experimental evidence from an in vitro model of epileptiform

discharge. Epilepsia 42S3, 2–4.

Badier, J., Chauvel, P., 1985. Electroencephalographic spiking activity, drug

levels, and seizure occurrence in epileptic patients. Ann. Neurol. 17 (6),

597–603.

Badier, J., Chauvel, P., 1995. Spatio-temporal characteristics of paroxysmal

interictal events in human temporal lobe epilepsy. J. Physiol. (Paris) 89

(4–6), 255–264.

Bagshaw, A.P., Aghakhani, Y., Benar, C.G., Kobayashi, E., Hawco, C.,

Dubeau, F., Pike, G.B., Gotman, J., 2004. EEG-fMRI of focal epileptic

spikes: analysis with multiple haemodynamic functions and comparison

with Gadolinium-enhanced MR angiograms. Hum. Brain Mapp. 22 (3),

179–192.

Bagshaw, A.P., Hawco, C., Benar, C.G., Kobayashi, E., Aghakhani, Y.,

Dubeau, F., Pike, G.B., Gotman, J., 2005. Analysis of the EEG-fMRI

response to prolonged bursts of interictal epileptiform activity. Neuro-

Image 24 (4), 1099–1112.

Benar, C.G., Gross, D.W., Wang, Y., Petre, V., Pike, G.B., Dubeau, F.,

Gotman, J., 2002. The BOLD response to interictal epileptiform

discharges. NeuroImage 17 (3), 1182–1192.

Bittar, R.G., Andermann, F., Olivier, A., Dubeau, F., Dumoulin, S.O., Pike,

G.B., Reutens, D.C., 1999. Interictal spikes increase cerebral glucose

metabolism and blood flow: a PET study. Epilepsia 40 (2), 170–178.

Boroojerdi, B., Diefenbach, K., Ferbert, A., 1996. Transcallosal inhibition

in cortical and subcortical cerebral vascular lesions. J. Neurol. Sci. 144,

160–170.

Boxerman, J.L., Bandettini, P.A., Kwong, K.K., Baker, J.R., Davis, T.L.,

Rosen, B.R., Weisskoff, R.M., 1995. The intravascular contribution to

fMRI signal change: Monte Carlo modeling and diffusion-weighted

studies in vivo. Magn. Reson. Med. 34 (1), 4–10 Jul.).

Breier, J.I., Mullani, N.A., Thomas, A.B., Wheless, J.W., Plenger, P.M.,

Gould, K.L., Papanicolaou, A., Willmore, L.J., 1997. Effects of

duration of epilepsy on the uncoupling of metabolism and blood flow

in complex partial seizures. Neurology 48 (4), 1047–1053.

Bruehl, C., Hagemann, G., Witte, O.W., 1998. Uncoupling of blood flow

and metabolism in focal epilepsy. Epilepsia 39 (12), 1235–1242.

Buckner, R.L., 1998. Event-related fMRI and the hemodynamic response.

Hum. Brain Mapp. 6 (5–6), 373–377.

Collins, D.L., Neelin, P., Peters, T.M., Evans, A.C., 1994. Automatic 3 D

intersubject registration of MR volumetric data in standardized

Talairach space. J. Comput. Assist. Tomogr. 18 (2), 192–205.

Cox, R.J., 1996. AFNI: software for analysis and visualization of functional

magnetic resonance neuroimages. Comput. Biomed. Res. 29, 162–173.

da Silva, E.A., Muller, R.A., Chugani, D.C., Shah, J., Shah, A., Watson, C.,

Chugani, H.T., 1999. Brain activation during intermittent photic

stimulation: a [15O]-water PET study on photosensitive epilepsy.

Epilepsia 40 (S4), 17–22.

Davis, T., Kwong, K., Weisskoff, R., Rosen, B., 1998. Calibrated functional

MRI: mapping the dynamics of oxidative metabolism. Proc. Natl. Acad.

Sci. U. S. A. 95, 1834–1839.

de Curtis, M., Avanzini, G., 2001. Interictal spikes in focal epileptogenesis.

Prog. Neurobiol. 63 (5), 541–567.

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215214

de Curtis, M., Manfridi, A., Biella, G., 1998. Activity-dependent pH shifts

and periodic recurrence of spontaneous interictal spikes in a model of

focal epileptogenesis. J. Neurosci. 18 (18), 7543–7551.

Deblieck, C., Pesenti, G., Scifo, P., Fazio, F., Bricolo, E., Russo, G.L.,

Scialfa, G., Cossu, M., Bottini, G., Paulesu, E., 2003. Preserved

functional competence of perilesional areas in drug-resistant epilepsy

with lesion in supplementary motor cortex: fMRI and neuropsycho-

logical observations. NeuroImage 20 (4), 2225–2234.

Devous, M.D., Leroy, R.F., Homan, R.W., 1990. Single photon emission

computed tomography in epilepsy. Semin. Nucl. Med. 20 (4), 325–341.

Diehl, B., Salek-Haddadi, A., Fish, D.R., Lemieux, L., 2003. Mapping of

spikes, slow waves, and motor tasks in a patient with malformation of

cortical development using simultaneous EEG and fMRI. Magn. Reson.

Imaging 21 (10), 1167–1173.

Duncan, J.S., 1997. Imaging and epilepsy. Brain 120 (Pt. 2), 339–377

Feb.).

Engel, J., Kuhl, D.E., Phelps, M.E., Mazziotta, J.C., 1982. Interictal

cerebral glucose metabolism in partial epilepsy and its relation to EEG

changes. Ann. Neurol. 12 (6), 510–517.

Engel, J., Lubens, P., Kuhl, D.E., Phelps, M.E., 1985. Local cerebral

metabolic rate for glucose during petit mal absences. Ann. Neurol. 17

(2), 121–128.

Evans, A.C., Collins, D.L., Mills, S.R., Brown, E.D., Kelly, R.L., Peters,

T.M., 1993. 3D statistical neuroanatomical models from 305 MRI

volumes. Proceedings of IEEE-Nuclear Science Symposium and

Medical Imaging Conference, pp. 1813–1817.

Ferbert, A., Rothwell, J.C., Day, B.L., Colebatch, J.G., Marsden, C.D.,

1992. Interhemispheric inhibition of the human motor cortex. J. Physiol.

453 (1), 525–546.

Fink, G.R., Pawlik, G., Stefan, H., Pietrzyk, U., Wienhard, K., Heiss, W.D.,

1996. Temporal lobe epilepsy: evidence for interictal uncoupling of

blood flow and glucose metabolism in temporomesial structures.

J. Neurol. Sci. 137 (1), 28–34.

Franck, G., Sadzot, B., Salmon, E., Depresseux, J.C., Grisar, T., Peters,

J.M., Guillaume, M., Quaglia, L., Delfiore, G., Lamotte, D., 1986.

Regional cerebral blood flow and metabolic rates in human focal

epilepsy and status epilepticus. Adv. Neurol. 44, 935–948.

Franck, G., Salmon, E., Sadzot, B., Maquet, P., 1989. Epilepsy: the use of

oxygen-15-labeled gases. Semin. Neurol. 9 (4), 307–316.

Gaillard, W.D., Fazilat, S., White, S., Malow, B., Sato, S., Reeves, P.,

Herscovitch, P., Theodore, W.H., 1995. Interictal metabolism and blood

flow are uncoupled in temporal lobe cortex of patients with complex

partial epilepsy. Neurology 45 (10), 1841–1847.

Gaillard, W.D., Zeffiro, T., Fazilat, S., DeCarli, C., Theodore, W.H., 1996.

Effect of valproate on cerebral metabolism and blood flow: an 18F-2-

deoxyglucose and 15O water positron emission tomography study.

Epilepsia 37 (6), 515–521.

Gerloff, C., Cohen, L.G., Floeter, M.K., Chen, R., Corwell, B., Hallett, M.,

1998. Inhibitory influence of the ipsilateral motor cortex on responses to

stimulation of the human cortex and pyramidal tract. J. Physiol. 510 (1),

249–259.

Gloor, P., 1978. Generalized epilepsy with bilateral synchronous spike and

wave discharge. New findings concerning its physiological mechanisms.

Electroencephalogr. Clin. Neurophysiol., Suppl. 34, 245–249.

Grisar, T., Lakaye, B., Thomas, E., Bettendorf, L., Minet, A., 1999. The

molecular neuron–glia couple and epileptogenesis. Adv. Neurol. 79,

591–602.

Grubb, R., Phelps, M., Eichling, J., 1974. The effects of vascular changes in

PaCO2on cerebral blood volume, blood flow and vascular mean transit

time. Stroke 5, 630–639.

Gutierrez-Delicado, E., Serratosa, J.M., 2004. Genetics of the epilepsies.

Curr. Opin. Neurol. 17 (2), 147–153 Apr.).

Haginoya, K., Munakata, M., Kato, R., Yokoyama, H., Ishizuka, M.,

Iinuma, K., 2002. Ictal cerebral haemodynamics of childhood epilepsy

measured with near-infrared spectrophotometry. Brain 125, 1960–1971.

Hamzei, F., Dettmers, C., Rzanny, R., Liepert, J., Buechel, C., Weiller, C.,

2002. Reduction of excitability (‘‘inhibition’’) in the ipsilateral primary

motor cortex is mirrored by fMRI signal decreases. NeuroImage 17,

490–496.

Hoge, R.D., Atkinson, J., Gill, B., Crelier, G., Marrett, S., Pike, G.,

1999. Investigation of BOLD signal dependence on CBF and CMRO2:

the deoxyhemoglobin dilution model. Magn. Reson. Med. 42 (5),

849–863.

Huang-Hellinger, F.R., Breiter, H.C., McCormack, G., Cohen, M.S.,

Kwong, K.K., Sutton, J.P., Savoy, R.L., Weiskoff, R.M., Davis, T.L.,

Baker, J.R., Belliveau, J.W., Rosen, B.R., 1995. Simultaneous func-

tional magnetic resonance imaging and electrophysiological recording.

Hum. Brain Mapp. 3, 13–23.

Huettel, S.A., McKeown, M.J., Song, A.W., Hart, S., Spencer, D.D.,

Allison, T., McCarthy, G., 2004. Linking hemodynamic and electro-

physiological measures of brain activity: evidence from functional MRI

and intracranial field potentials. Cereb. Cortex 14 (2), 165–173.

Ives, J.R., Warach, S., Schmitt, F., Edelman, R.R., Schomer, D.L., 1993.

Monitoring the patient’s EEG during echo planar MRI. Electro-

encephalogr. Clin. Neurophysiol. 87 (6), 417–420.

Jager, L., Werhahn, K.J., Hoffmann, A., Berthold, S., Scholz, V., Weber, J.,

Noachtar, S., Reiser, M., 2002. Focal epileptiform activity in the brain:

detection with spike-related functional MR imaging—Preliminary

results. Radiology 223 (3), 860–869.

Janszky, J., Fogarasi, A., Jokeit, H., Schulz, R., Hoppe, M., Ebner, A.,

2001. Spatiotemporal relationship between seizure activity and interictal

spikes in temporal lobe epilepsy. Epilepsy Res. 47 (3), 179–188.

Jensen, M.S., Yaari, Y., 1997. Role of intrinsic burst firing, potassium

accumulation, and electrical coupling in the elevated potassium model

of hippocampal epilepsy. J. Neurophysiol. 77 (3), 1224–1233.

Josephs, O., Turner, R., Friston, K., 1997. Event-related fMRI. Hum. Brain

Mapp. 5, 1–7.

Kang, J.K., Benar, C.G., Al-Asmi, A., Khani, Y.A., Pike, G.B., Dubeau, F.,

Gotman, J., 2003. Using patient-specific hemodynamic response

functions in combined EEG-fMRI studies in epilepsy. NeuroImage 20

(2), 1162–1170.

Kanner, A.M., Kaydanova, Y., de Toledo-Morrell, L., Morrell, F., Smith,

M.C., Bergen, D., Pierre-Louis, S.J., Ristanovic, R., 1995. Tailored

anterior temporal lobectomy. Relation between extent of resection of

mesial structures and postsurgical seizure outcome. Arch. Neurol. 52

(2), 173–178.

Kapucu, L.O., Serdaroglu, A., Okuyaz, C., Kose, G., Gucuyener, K.,

2003. Brain single photon emission computed tomographic evaluation

of patients with childhood absence epilepsy. J. Child. Neurol. 18 (8),

542–548.

Kastrup, A., Krueger, G., Neumann-Haefelin, T., Glover, G., Moseley,

M.E., 2002. Changes of cerebral blood flow, oxygenation and oxidative

metabolism during graded motor activation. NeuroImage 15, 74–82.

Kelly, P.A., McCulloch, J., 1983. The effects of the GABAergic agonist

muscimol upon the relationship between local cerebral blood flow and

glucose utilization. Brain Res. 258 (2), 338–342.

Kim, S.-K., Lee, D.S., Lee, S.K., Kim, Y.K., Kang, K.W., Chung, C.K.,

Chung, J.-K., Myung, C.L., 2001. Diagnostic performance of

[18F]FDG-PET and ictal [99mTc]-HMPAO SPECT in occipital lobe

epilepsy. Epilepsia 42 (12), 1531–1540.

Krakow, K., Woermann, F.G., Symms, M.R., Allen, P.J., Lemieux, L.,

Barker, G.J., Duncan, J.S., Fish, D.R., 1999. EEG-triggered functional

MRI of interictal epileptiform activity in patients with partial seizures.

Brain 122, 1679–1688.

Kreisman, N.R., LaManna, J.C., Rosenthal, M., Sick, T.J., 1981. Oxidative

metabolic responses with recurrent seizures in rat cerebral cortex: role

of systemic factors. Brain Res. 218 (1–2), 175–188.

Kreisman, N.R., Rosenthal, M., Sick, T.J., LaManna, J.C., 1983. Oxidative

metabolic responses during recurrent seizures are independent of

convulsant, anesthetic, or species. Neurology 33 (7), 861–867.

Kuhl, D.E., Engel, J., Phelps, M.E., Selin, C., 1980. Epileptic patterns of

local cerebral metabolism and perfusion in humans determined by

emission computed tomography of 18 FDG and 13N-H3. Ann. Neurol.

8 (4), 348–360.

B. Stefanovic et al. / NeuroImage 28 (2005) 205–215 215

Lazeyras, F., Blanke, O., Perrig, S., Zimine, I., Golay, X., Delavelle, J.,

Michel, C.M., de Tribolet, N., Villemure, J.G., Seeck, M., 2000. EEG-

triggered functional MRI in patients with pharmacoresistant epilepsy.

J. Magn. Reson. Imaging 12 (1), 177–185.

Lee, B.I., Markand, O.N., Siddiqui, A.R., Park, H.M., Mock, B., Wellman,

H.H., Worth, R.M., Edwards, M.K., 1986. Single photon emission

computed tomography (SPECT) brain imaging using N,N,NV-trimethyl-

NV-(2 hydroxy-3-methyl-5-123I-iodobenzyl)-1,3-propanediamine 2 HCl

(HIPDM): intractable complex partial seizures. Neurology 36 (11),

1471–1477.

Leiderman, D.B., Balish, M., Bromfield, E.B., Theodore, W.H., 1991.

Effect of valproate on human cerebral glucose metabolism. Epilepsia 32

(3), 417–422.

Lemieux, L., Salek-Haddadi, A., Josephs, O., Allen, P., Toms, N., Scott, C.,

Krakow, K., Turner, R., Fish, D.R., 2001. Event-related fMRI with

simultaneous and continuous EEG: description of the method and initial

case report. NeuroImage 14 (3), 780–787.

Liepert, J., Dettmers, C., Terborg, C., Weiller, C., 2001. Inhibition of

ipsilateral motor cortex during phasic generation of low force. Clin.

Neurophysiol. 112, 114–121.

Logothetis, N., Pauls, J., Augath, M., Trinath, T., Oeltermann, A., 2001.

Neurophysiological investigation of the basis of the fMRI signal. Nature

412, 150–157.

Loscher, W., 1979. GABA in plasma and cerebrospinal fluid of different

species. Effects of gamma-acetylenic GABA, gamma-vinyl GABA and

sodium valproate. J. Neurochem. 32 (5), 1587–1591.

Loscher, W., 1981. Valproate induced changes in GABA metabolism at the

subcellular level. Biochem. Pharmacol. 30 (11), 1364–1366.

Macdonald, R.L., Gallagher, M.J., Feng, H.J., Kang, J., 2004.

GABA(A) receptor epilepsy mutations. Biochem. Pharmacol. 68

(8), 1497–1506.

Matsumoto, H., Ajmone Marsan, C., 1964a. Cortical cellular phenomena in

experimental epilepsy: ictal manifestations. Exp. Neurol. 25, 305–326.

Matsumoto, H., Ajmone Marsan, C., 1964b. Cortical cellular phenomena

in experimental epilepsy: interictal manifestations. Exp. Neurol. 80,

286–304.

McKhann, G.M., Schoenfeld-McNeill, J., Born, D.E., Haglund, M.M.,

Ojemann, G.A., 2000. Intraoperative hippocampal electrocorticography

to predict the extent of hippocampal resection in temporal lobe epilepsy

surgery. J. Neurosurg. 93 (1), 44–52.

Metrakos, K., Metrakos, J.D., 1961. Genetics of convulsive disorders:

II. Genetic and electroencephalographic studies in centrencephalic

epilepsy. Neurology 11, 474–483.

Netz, J., Ziemann, U., Hoemberg, V., 1995. Hemispheric asymmetry of

transcallosal inhibition in man. Exp. Brain Res. 104, 527–533.

Ochs, R.F., Gloor, P., Tyler, J.L., Wolfson, T., Worsley, K., Andermann, F.,

Diksic, M., Meyer, E., Evans, A., 1987. Effect of generalized spike-and-

wave discharge on glucose metabolism measured by positron emission

tomography. Ann. Neurol. 21 (5), 458–464.

Palfreyman, M.G., Huot, S., Grove, J., 1983. Total GABA and homo-

carnosine in CSF as indices of brain GABA concentrations. Neurosci.

Lett. 35 (2), 161–166.

Patel, M.R., Blum, A., Pearlman, J.D., Yousuf, N., Ives, J.R., Saeteng, S.,

Schomer, D.L., Edelman, R.R., 1999. Echo-planar functional MR

imaging of epilepsy with concurrent EEG monitoring. Am. J. Neuro-

radiol. 20 (10), 1916–1919.

Penfield, W., Jasper, H., 1954. Epilepsy and the Functional Anatomy of the

Human Brain. Little, Brown and Company, Boston, MA.

Petroff, O.A., Rothman, D.L., Behar, K.L., Collins, T.L., Mattson, R.H.,

1996. Human brain GABA levels rise rapidly after initiation of

vigabatrin therapy. Neurology 47 (6), 1567–1571.

Press, W.H., Teukolsky, S.A., Vetterling, W.T., Flannery, B.P., 1992.

Numerical Recipes in C: The Art of Scientific Computing, 2nd edition.

Cambridge University Press, New York, pp. 666–670 (Chapt. 15).

Prevett, M.C., Duncan, J.S., Jones, T., Fish, D.R., Brooks, D.J., 1995.

Demonstration of thalamic activation during typical absence seizures

using H2(15)O and PET. Neurology 45 (7), 1396–1402.

Rosenow, F., Luders, H., 2001. Presurgical evaluation of epilepsy. Epilepsia

42 (12), 1515–1522.

Salek-Haddadi, A., Friston, K.J., Lemieux, L., Fish, D.R., 2003a. Studying

spontaneous EEG activity with fMRI. Brain Res. Rev. 43 (1), 110–133.

Salek-Haddadi, A., Lemieux, L., Merschhemke, M., Friston, K.J., Duncan,

J.S., Fish, D.R., 2003b. Functional magnetic resonance imaging of

human absence seizures. Ann. Neurol. 53 (5), 663–667.

Savic, I., Osterman, Y., Helms, G., 2004. MRS shows syndrome diffe-

rentiated metabolite changes in human-generalized epilepsies. Neuro-

Image 21 (1), 163–172.

Seeck, M., Lazeyras, F., Michel, C.M., Blanke, O., Gericke, C.A., Ives, J.,

Delavelle, J., Golay, X., Haenggeli, C.A., de Tribolet, N., Landis, T.,

1998. Non-invasive epileptic focus localization using EEG-triggered

functional MRI and electromagnetic tomography. Electroencephalogr.

Clin. Neurophysiol. 106, 508–512.

Spanaki, M.V., Siegel, H., Kopylev, L., Fazilat, S., Dean, A., Liow, K., Ben-

Menachem, E., Gaillard, W.D., Theodore, W.H., 1999. The effect of

vigabatrin (gamma-vinyl GABA) on cerebral blood flow and metabo-

lism. Neurology 53 (7), 1518–1522.

Sperling, M.R., Skolnick, B.E., 1995. Cerebral blood flow during spike-

wave discharges. Epilepsia 36 (2), 156–163.

Stefanovic, B., Warnking, J.M., Pike, G.B., 2004. Hemodynamic and

metabolic responses to neuronal inhibition. NeuroImage 22 (2),

771–778.

Szaflarski, J.P., Holland, S.K., Schmithorst, V.J., Dunn, R.S., Privitera,

M.D., 2004. High-resolution functional MRI at 3 T in healthy and

epilepsy subjects: hippocampal activation with picture encoding task.

Epilepsy Behav. 5 (2), 244–252.

Tassinari, C.A., Cincotta, M., Zaccara, G., Michelucci, R., 2003. Trans-

cranial magnetic stimulation and epilepsy. Clin. Neurophysiol. 114 (5),

777–798.

Theodore, W.H., 1988. Antiepileptic drugs and cerebral glucose metabo-

lism. Epilepsia 29S2, S48–S55.

Theodore, W.H., Brooks, R., Margolin, R., Patronas, N., Sato, S., Porter,

R.J., Mansi, L., Bairamian, D., DiChiro, G., 1985. Positron emission

tomography in generalized seizures. Neurology 35 (5), 684–690.

Theodore, W.H., Bromfield, E., Onorati, L., 1989. The effect of

carbamazepine on cerebral glucose metabolism. Ann. Neurol. 25 (5),

516–520.

Warach, S., Ives, J.R., Schlaug, G., Patel, M.R., Darby, D.G., Thangaraj, V.,

Edelman, R.R., Schomer, D.L., 1996. EEG-triggered echo-planar

functional MRI in epilepsy. Neurology 47 (1), 89–93.

Warnking, J.M., Pike, G.B., 2004. Bandwidth-modulated adiabatic RF

pulses for uniform selective saturation and inversion. Magn. Reson.

Med. 52 (5), 1190–1199.

Wong, E., Buxton, R., Frank, L., 1997. Implementation of quantitative

perfusion imaging techniques for functional brain mapping using pulsed

arterial spin labeling. NMR Biomed. 10, 237–249.

Worsley, K., Liao, C., Aston, J., Petre, V., Duncan, G., Morales, F., Evans,

A., 2002. A general statistical analysis for fMRI data. NeuroImage 15,

1–15.