Neuroradiolog y • Original Research

Phal et al.
MRI of Epilepsy

Neuroradiology
Original Research
Downloaded from www.ajronline.org by 103.213.128.111 on 02/26/19 from IP address 103.213.128.111. Copyright ARRS. For personal use only; all rights reserved

Qualitative Comparison of 3-T

and 1.5-T MRI in the Evaluation
of Epilepsy
Pramit M. Phal1 Objective. MRI at 3 T, which has a higher signal-to-noise ratio than 1.5-T MRI, is
Alexander Usmanov 1 potentially more sensitive and specific at delineating epileptogenic lesions and may influence
Gary M. Nesbit 1 management of refractory epilepsy. The purposes of the current study were to compare image
James C. Anderson1 quality of 3-T MRI with that of 1.5-T MRI in the evaluation of epilepsy and, in cases of focal
David Spencer 2 epilepsy, to compare the two field strengths in terms of lesion detection and characterization.
MATERIALS and Methods. Retrospective review was performed on 50 sets of
Paul Wang1
MR images of 25 patients who underwent both 3-T and 1.5-T brain imaging with a dedicated
Jonathan A. Helwig1 epilepsy protocol, including fast spin-echo T2-weighted, coronal FLAIR, coronal fast
Colin Roberts 2 multiplanar inversion recovery, and 3D spoiled gradient-recalled echo pulse sequences.
Bronwyn E. Hamilton1 Parameters assessed were distortion and artifact, lesion conspicuity, gray–white matter
differentiation, and motion. Each pulse sequence was graded on a 4-point scale. Reviewers
Phal PM, Usmanov A, Nesbit GM, et al.
performed qualitative assessments of the site of abnormality and the most likely diagnosis.
RESULTS. MRI at 3 T outperformed MRI at 1.5 T in all four parameters and was
statistically superior (p < 0.05) to 1.5-T MRI in all categories except motion. On 3-T MRI,
lesions were detected in 65 of 74 cases compared with 55 of 74 cases at 1.5 T (p = 0.0364),
and lesions were accurately characterized in 63 of 74 cases compared with 51 of 74 cases at
1.5 T (p = 0.0194). The odds ratios showed identification of a focal epileptogenic lesion with
3-T MRI 2.57 times as likely as identification with 1.5-T MRI and accurate characterization
of lesions 2.66 times as likely as characterization with 1.5-T MRI.
CONCLUSION. In evaluation of epilepsy, MRI at 3 T performed better than 1.5-T MRI
in image quality, detection of structural lesions, and characterization of lesions. High-field-
strength imaging should be considered for patients with intractable epilepsy and normal or
equivocal findings on 1.5-T MRI.

E
pilepsy is a disease with serious
consequences for patients and
society [1–3]. For patients with
Keywords: 3-T MRI, focal epilepsy, medically refractory
partial complex epilepsy, identi­
epilepsy, MRI
fying a focal structural brain abnormality
DOI:10.2214/AJR.07.3933 offers the best potential for surgical cure and
improvement in quality of life. For patients
Received March 3, 2008; accepted after revision with medically refractory focal epilepsy, sur­
April 13, 2008.
gery not only is the single remaining treat­
1
Division of Neuroradiology, Oregon Health & Science ment option but also offers the best chance
University, 3181 SW Sam Jackson Park Rd,, Mail Code for a permanent cure and is the most cost-
CR 135, Portland, OR 97239. Address correspondence effective approach in the long term [4, 5].
to B. E. Hamilton (hamiltob@ohsu.edu).
The current barrier to surgery for many
2
Division of Neurology, Oregon Health & Science patients with medically refractory partial
University, Portland, OR. complex epilepsy is lack of identification of
an abnormality on images. MRI is key to
AJR 2008; 191:890–895
surgical success because it enables accurate
0361–803X/08/1913–890 anatomic identification of the epileptogenic
focus, which is critical for preoperative
© American Roentgen Ray Society planning and localization [6, 7]. Assessment
is ideally performed at dedicated epilepsy
centers with close collaboration among neu­
rologists, neurosurgeons, and radiolo­gists.
Safe surgery requires careful analysis of
structural brain lesions with alignment of the
clinical and imaging evidence. MRI plays a
key role in diagnosis and localiza­tion. Many
patients presenting to our epi­lepsy center
have localized syndromes that raise clinical
suspicion of the presence of a focal struc­
tural abnormality, yet in many instances, a
lesion has not been localized at previous
imaging evaluation. High-field-strength MRI
has potential for improving epilepsy evalu­
ation because of the greater signal-to-noise
ratio of 3-T MRI compared with 1.5-T MRI.
The goal of our study was to assess the
diagnostic value of 3-T compared with 1.5-T
whole-brain MRI in the evaluation of epi­
lepsy. We selected for review all patients

890 AJR:191, September 2008


who underwent both 3-T and 1.5-T whole-
brain MRI for epilepsy regardless of the
reason for repeated imaging. We evaluated the
pro­portions of correct detection of structural
lesions, observer-assessed lesion conspicuity,
Downloaded from www.ajronline.org by 103.213.128.111 on 02/26/19 from IP address 103.213.128.111. Copyright ARRS. For personal use only; all rights reserved
normal gray–white matter tissue contrast, and
imaging artifacts in a group of epilepsy pat­
ients who had undergone consecutive 1.5- and
3-T MRI examinations at our institution.
Materials and Methods
We reviewed images from all available 3-T
MRI examinations performed from March 2003
to December 2005 on all patients referred
because of epilepsy. Inclusion criteria were that
all patients had undergone both 1.5-T and 3-T
whole-brain MRI according to our epilepsy
protocol between January 2000 and December
2005. Indications for repeated evaluation at 3 T
varied. Although some examinations were per­
formed because of prev­iously normal or equivocal
results at 1.5 T, many others were performed for
lesional follow-up and surgical planning and
because of scheduling con­straints related to
availability of the MRI unit. Studies without
directly comparable high-resolution sequences
were excluded, so a total of 25 patients who
underwent 50 MRI examinations were included
in the review. This retrospective study was
approved by our institutional review board with
waiver of informed consent.
The reference standard for lesion localization in
the 19 patients with partial complex epilepsy was
surgical confirmation in 12 cases and electro­
encephalographic or PET localization in con­
junction with clear clinical signs in seven cases.
The other six patients did not have a focal epilepsy
syndrome, and their cases were used only for the
qualitative assessment portion of the analysis.
MRI
Both 1.5-T and 3-T MR images of the 25
patients were reviewed independently by four
experienced neuroradiologists. The images were
assessed digitally with a commercially available
PACS workstation (Impax version 4.5, Agfa)
with real-time multiplanar reformation capa­
bilities available to all reviewers. The multiplanar
reformation function operates with a localization
marker on both the source and the reformatted
images. This feature was particularly helpful in
assessment of the 3D T1-weighted spoiled
gradient-recalled echo (SPGR) images with
nearly isovoxel resolution. With this tool, the
radiologist was able to assess areas suggestive of
cortical thickening in directly orthogonal or
perpendicular planes to rule out artifacts related
to in-plane cortex. Reviewers were blinded to
AJR:191, September 2008 891
MRI of Epilepsy
clinical results, including data on seizure signs,
electro­encephalographic findings, and other
forms of localization.
A six-channel sensitivity-encoding head coil
was used on both clinical 3-T units (Achieva,
Philips Healthcare) for all whole-brain epilepsy
imaging. Our 1.5-T MRI units (Signa Horizon and
Signa LX, GE Healthcare) had a transmit–receive
single-channel head coil for whole-brain imaging.
Parallel-processing head coils were impractical on
our two 1.5-T units for several reasons but
primarily owing to degradation in image quality
from inadequate signal-to-noise ratio. Identical
imaging parameters therefore were not possible.
Directly comparable sequences (those of the same
sequence type, plane, and approximate slice
thickness) used for our epilepsy protocol on the
3-T and 1.5-T MRI units were reviewed. At the
time of this study, our whole-brain epilepsy
protocol on all units included the following
sequence parameters.
The 1.5-T protocol consisted of one 3D and
three 2D sequences. The 3D images were
obtained with a coronal T1-weighted SPGR
sequence (TR/TE, 24/9.2; acquisition matrix,
256 × 256; field of view, 230 mm 2 ; flip angle,
25°; slice thickness, 1.5 mm with no space). The
first 2D acquisition was an axial fast spin-echo
T2-weighted sequence (5,000/96.1; acquisition
matrix, 256 × 256; field of view, 230 mm 2 ; flip
angle, 90°; slice thickness, 4.0–5.0 mm with
1.0-mm space). Two-dimensional fast multiplanar
inversion recovery (4,500/14; inversion time,
300 seconds; acquisition matrix, 256 × 256; field
of view, 180–220 mm; slice thickness, 3.0 mm
with no space) and coronal FLAIR (8,802/133;
inversion time, 2,200 milliseconds; acquisition
matrix, 256 × 256; field of view, 220–240 mm;
slice thickness, 5.0 mm with 1.0-mm space)
sequences also were performed.
The 3-T protocol also consisted of one 3D and
three 2D sequences. The 3D images were obtained
with a coronal T1-weighted SPGR sequence (30/6;
acquisition matrix, 256 × 256; field of view, 230
mm; flip angle, 45°; slice thickness, 1.2 mm with
no space). The first 2D acquisition was an axial
turbo spin-echo T2-weighted sequence (3,000/90;
acqui­sition matrix, 256 × 256; field of view, 230
mm; flip angle, 90°; slice thickness, 4.0–5.0 mm
with 1.0-mm space). Two-dimensional fast multi­
planar inversion recovery (3,975/20; inversion
time, 250 seconds; acquisition matrix, 256 × 256;
field of view, 180–220 mm; slice thickness, 2.0 mm
thick with 0.2-mm space) and coronal FLAIR
(11,004/ 120; inversion time, 2,800 milliseconds;
acquisi­tion matrix, 256 × 256; field of view,
220–240 mm; slice thickness, 4.0 mm with 1.0-mm
space) sequences also were performed.
Image Review
Four neuroradiologists experienced in inter­
preting epilepsy studies were asked to independ­
ently review the images from the 1.5- and 3-T
studies. The viewing order was random, and to
allow them the opportunity for direct comparison,
reviewers were not blinded in regard to viewing
both studies at the same time. Reviewers were
asked to rate the 1.5- and 3-T image sets separately
for the four following features: lesion conspicuity,
defined as the ease with which the suspected
epileptogenic focus was visible, with a specific
diagnosis when possible; normal tissue contrast
between gray and white matter; technical artifacts
resulting in image degradation; and artifacts
related to patient motion. All reviewers were
blinded to clinical findings, final diagnosis, and
other reviewers’ interpretations. All features were
rated on a 4-point scale (1, worst; 4, best) for lesion
conspicuity and tissue contrast (1, worst artifacts;
4, clinically insignificant or no artifacts) for image
degradation due to technical factors for both
overall imaging artifacts, such as phase and
susceptibility artifacts, and motion.
Statistical Evaluation
Analysis of variance was used to assess dif­
ferences in the reported scores of lesion char­
acterization, tissue contrast, and technical and
motion artifacts. Differences in reported identi­
fication also were compared because in some
cases, anatomic abnormalities were visible only at
3 T. Individual scores were used as the response
variable, and p = 0.05 was considered significant.
Logistic regression was used to determine the
diagnostic accuracy of 3-T com­pared with 1.5-T
MRI through the use of two models fitted for
lesion characterization, tissue contrast, and
technical and motion artifacts as responses. A
value of p < 0.05 was considered significant.
Intraclass correlation is a measure of interrater
reliability for two or more reviewers, and the
significance of this value can be interpreted in a
manner similar to that for kappa statistics. The
95% CI for intraclass correlation was used to
assess the reliability of the four independent
reviewers’ scores.
Results
Patients and Pathologic Findings
A total of 13 pediatric and 12 adult patients
(mean age, 24 years; range, 10 months–70
years) were included in the review. The mean
time between 1.5- and 3-T MRI was 1.3
years (range, 0.2–5 years). Diagnoses in the
19 cases of focal epilepsy were established
histologically after surgical resection in nine
cases and on the basis of characteristic clinical
 Phal et al.

semiologic and imaging findings in the other
10 cases. The focal epilepsy syndromes
diagnosed were cortical dysplasia in eight
cases; two cases each of dysembryoplastic
neuroepithelial tumor, mesial temporal
Downloaded from www.ajronline.org by 103.213.128.111 on 02/26/19 from IP address 103.213.128.111. Copyright ARRS. For personal use only; all rights reserved
Correct lesion identification (separate from
the quality analysis) was higher at 3 T than at
1.5 T with correct identification of the
structural lesion in 65 of 74 (88%) compared
with 55 of 74 (74%) individual interpretations
ification and characterization at 3 T and 1.5 T
were 2.57 and 2.66, respectively (Table 4). An
interesting finding was that imaging artifacts
were less troublesome at 3 T than at 1.5 T (p =
0.01). Although a trend toward greater mo­

sclerosis, hypothalamic hamartoma, caver­
nous malformation, and ganglio­glioma; and
one case of oligodendroglial hyperplasia
(Figs. 1–3). Because they had generalized
epilepsy syndromes without a structural
lesion identified, the other six patients were
included only in the image quality portion of
the assessment (normal tissue contrast and
technical or motion artifacts).
(Table 2). Two of the 76 interpretations in the
cases of the 19 focal epilepsy patients were
excluded because of missing data. Lesion
characterization at 3 T and 1.5 T was compared
only for the 19 of 26 patients with focal
epilepsy (Table 3). Lesion characterization
(p = 0.0095) and tissue contrast (p = 0.0292)
were consistently rated higher at 3 T than at
1.5 T. Odds ratios for correct lesion ident­
tion artifacts was seen at 3 T, this difference
was not statistically significant (p = 0.136).
Intraclass correlation yielded moderate reli­
ability among the four reviewers as a group
(0.562) with a 95% CI of 0.466–0.643.
Clinical Outcome
Twelve of the 19 patients with focal epi­
lepsy underwent surgical resection of the

Image Assessment
For each of the four quality parameters TABLE 1:  Mean Scores in Imaging Quality Assessment
assessed, mean composite scores were higher Parameter Mean Score at 1.5 T Mean Score at 3 T F p
for 3-T than for 1.5-T MRI (Table 1). There
Technical artifacts 2.46 (0.05) 3.16 (0.05) 8.70 0.0100
were statistically significant differences in
three of four parameters assessed: lesion Lesion conspicuity 2.29 (0.05) 2.95 (0.05) 8.84 0.0095
characterization, tissue contrast, and imaging Gray–white matter contrast 2.24 (0.04) 2.73 (0.04) 5.81 0.0292
artifacts other than those related to patient Motion-related artifacts 2.45 (0.03) 2.88 (0.03) 2.48 0.1360
motion. Motion artifacts were com­parable at
Note—Values in parentheses are standard error. Analysis of variance was used to determine the components
3 T and 1.5 T without a statistically signi­ of total variance, calculated as the F ratio, that is, factor variance (between groups) divided by error variance
ficant difference. (within group).

Fig. 1—17-year-old girl with intractable nocturnal

seizures starting at age 2 years.
A, Coronal FLAIR 1.5-T MR image shows questionable
curvilinear focus of juxtacortical high signal
intensity (arrows) in left occipital lobe. Abnormal
signal intensity was missed at first review of images.
B, Coronal FLAIR 3-T MR image shows curvilinear
band of high signal intensity (arrows) in left occipital
juxtacortical white matter without apparent mass
effect. Focus was surgically resected, and histologic
finding was focal cortical dysplasia with balloon cell
features (Taylor’s type).
A B

Fig. 2—3-year-old boy with intractable left temporal

lobe epilepsy.
A, Coronal FLAIR 1.5-T MR image shows questionable
area of subtly decreased signal intensity and size of
anterior left hippocampus (arrow).
B, Coronal FLAIR 3-T MR image shows left
hippocampus smaller than in A with associated
high signal intensity (arrow) and loss of internal
architecture consistent with mesial temporal
sclerosis. Surgical resection at age 6 years showed
histologic findings confirming presence of mesial
temporal sclerosis.
A B

892 AJR:191, September 2008

 MRI of Epilepsy

Fig. 3—3-year-old boy with intractable seizures.

A, Coronal thin-slice T1-weighted spoiled gradient-
recalled echo MR image suggests thickening (arrows)
of right frontal cortex.
B, Companion 3-T coronal T1-weighted spoiled
gradient-recalled echo MR image shows thickened
indistinct cortex (arrows) in right frontal lobe. Gray–
Downloaded from www.ajronline.org by 103.213.128.111 on 02/26/19 from IP address 103.213.128.111. Copyright ARRS. For personal use only; all rights reserved

white matter contrast is better than in A.

A B
C, Coronal 1.5-T STIR MR image suggests decreased
arborization of normally T2-weighted hypointense
white matter in right frontal lobe. Suggestion of
thickened-appearing cortex (solid arrows) can easily
be interpreted as volume averaging. Normal white-
matter arborization (open arrow) is evident in left
frontal lobe at site of thin cortical ribbon.
D, Companion 3-T STIR MR image shows thickening
and indistinctness of right frontal gray–white matter
junction (solid arrows) better than does C. Normal
white-matter arborization (open arrow) is present
where cortical ribbon shows normal thickness.
After surgical resection of abnormal-appearing
tissue at age 3 years, histologic finding was cortical
glioneuronal dysplasia.

C D

epileptogenic focus. Seven of the 12 surgi­
cally treated patients had complete resolution
of seizures, three had clinically moderate
improvement, and two had no significant
improvement in seizure frequency or severity
during the early clinical follow-up period
(mean, 571 days).
Discussion
It is estimated that 60% of epilepsy patients
have a focal syndrome and that in 25% of
those cases, epilepsy is not controlled with
anti­convulsive medication [8]. Medically
refractory epilepsy can be debil­itating and is
associated with considerable morbidity [9].
For patients with an epi­leptogenic focus identi­
fiable at imaging, surgery offers the potential
for long-term cure [1–3]. When clinical signs
of seizure and electroencephalographic and
PET findings are concordant with anatomic
localization on MRI, surgery can be safely
undertaken in most patients [9]. Studies
of dedicated high-resolution 1.5-T MRI in
the evaluation of epilepsy show utility in
localization and characterization of structural
abnormalities [1].
The results of our study support the clinical
supposition that use of 3-T MRI increases the
rates of lesion detection and accurate char­
acterization of lesions. MRI at 3 T also yields
better contrast resolution of the gray–white
matter junction, a finding particularly relevant
for detection of subtle focal dysplasia of the
cortex. Data from our odds ratio comparison
imply that a 3-T MRI examination is 2.57 as
likely as a 1.5-T examination to depict a

TABLE 2:  Lesion Identification TABLE 3:  Lesion Characterization

Result 1.5 T 3T Result 1.5 T 3T
No. of lesions identified 55/74 65/74 No. of lesions well characterized 51/74 63/74
No. of lesions not identified 19/74 9/74 No. of lesions poorly characterized 23/74 11/74
Percentage of lesions identified 74 88 Percentage of lesions well characterized 69 85
Note—Composite numbers (data from all four reviewers) of structural abnormali- Note—Lesion characterization reflects the reviewer’s level of confidence in
ties correctly identified on MRI for the 19 patients with focal epilepsy. Two of 76 accurately assessing a structural abnormality on the basis of the intrinsic MRI
interpretations were excluded because of missing data. Such lesions were signal characteristics and inherent tissue contrast of the lesion relative to
confirmed surgically or at retrospective review of all available surgical histologic, surrounding normal brain parenchyma. Two of 76 interpretations were excluded
correlative imaging, semiologic, and clinical data. because of missing data. High confidence levels (scores of 3 and 4) reflect
well-characterized lesions; scores of 1 and 2 indicate poor characterization.

AJR:191, September 2008 893


TABLE 4: Comparison of Lesion Identification and Characterization at 1.5 T
and 3 T
Result p (3 T vs 1.5 T)
Lesion identification 0.0364
Lesion characterization 0.0194
Downloaded from www.ajronline.org by 103.213.128.111 on 02/26/19 from IP address 103.213.128.111. Copyright ARRS. For personal use only; all rights reserved
Note—Odds ratios were calculated for the 19 patients with structurally identified lesions. The data imply that a
3-T study is 2.574 times as likely as a 1.5-T study to depict the abnormality. Similarly, the rate of accurate lesion
characterization (defined as lesion contrast relative to surrounding normal brain) with 3-T MRI is 2.664 times
as high as that with 1.5-T MRI.
structural abnormality and that correct char­
acterization of the abnormality is 2.66 times
as likely on 3-T studies as it is on 1.5-T studies,
presumably contributing to a more accurate
diagnosis. MRI at 3-T with its intrinsically
greater signal-to-noise ratio combined with
advances in parallel processing has consider­
able diagnostic value. The ability to produce
high-­resolution thin-slice whole-brain images
in a practical examination time (< 1 hour) is a
strong advantage on 3-T MRI because it
facilitates detailed anatomic evaluation with
minimal artifacts.
Image acquisition at 3 T combined with
parallel processing makes 3D volume acqui­
sition at nearly isovoxel resolution through
the entire brain a practical reality. Although
technically feasible on our clinical 1.5-T MRI
units, this sequence was less desirable from a
diagnostic standpoint owing to degradation
in image quality due to increased noise and
time constraints. High-resolution 3D volume
isovoxel acquisitions facilitate multiplanar
reformation in any plane, which is important
for accurate lesion differentiation from nor­
mal gray–white matter structures. Reformat­
ting a gyrus with in-plane orient­ation into a
perpendicular orthogonal plane can be criti­
cal for avoiding the volume-averaging effects
that result in overcalling gray–white matter
thickening or indistinct­ness. These sequences
are also highly desirable for neurosurgeons
in preoperative planning.
Since the completion of our study, high-
resolution 3D T2-weighted and FLAIR tech­
niques have become commercially available
for our 3-T MRI units. Although we aim to
explore the advantages of such techniques at
3 T, they are not practically feasible at 1.5 T
because of time constraints; therefore, the
techniques are not directly comparable.
Our results are concordant with findings
reported for 3-T MRI with eight-channel
phased-array surface coils in the evaluation
of focal epilepsy [10]. The major gains in
diagnosis in our study not surprisingly were
related to detection of cortical malformations,
894 AJR:191, September 2008
Phal et al.
Odds Ratio (3 T vs 1.5 T) 95% CI for Odds Ratio
2.574 1.062–6.240
2.664 1.172–6.055
because many of our patients are referred for
follow-up MRI because of failures of imaging
localization with 1.5-T MRI. These
abnormalities are often subtle; thin slices are
needed to avoid partial volume effects and to
detect subtle areas of gray–white matter
blurring and indistinctness [11, 12]. The
improved gray–white matter contrast at 3 T
in our study highlights the importance of this
factor. MRI at 3 T was excellent for depiction
of the gray–white matter junction (tissue
contrast) in our study, showing statistically
significant improvement over 1.5-T MRI.
The limitations of high-field-strength im­
aging are well known, including a propen­
sity to certain imaging artifacts, such as
sus­ceptibility and a perceived sensitivity to
motion. Although usually undesirable, great­er
susceptibility effects at 3 T can be ad­van­
tageous, as when detection of abnormal ves­
sels or previous hemorrhage is relevant. Detec­
tion of two cavernous malformations in our
study was improved by susceptibility effects,
which were greater at 3 T than at 1.5 T.
Disadvantages of imaging at 3 T include
longer T1, increased acoustic noise, greater
power deposition, and greater device incom­
patibility [13–16]. Although motion artifacts
were found to be similar at 1.5 and 3 T, other
imaging artifacts were notably fewer at 3 T
in our study. We did not anticipate this
finding, which might have been related in
part to the relatively young age of the patient
population, who have minimal or no sus­
ceptibility problems related to previous
intracranial surgery or hemorrhage.
Our study had several limitations. The
retrospective nature of the review, the
indications for a second MRI examination at
3 T, and the need to exclude patients without
directly comparable sequences may have
introduced selection bias. Epilepsy patients
undergo 3-T MRI at our institution for
various reasons, including normal or equi­
vocal findings on 1.5-T MRI, clarification of
lesion characterization, surgical planning,
and scheduling constraints. However, be­
cause normal or equivocal findings on 1.5-T
MRI should theoretically increase the
likelihood of normal findings on 3-T MRI
(implying a lower pretest probability of
disease), repetition of imaging at 3 T should
adversely affect rather than improve the
likelihood of identification of a structural
lesion. By contrast, we found more than
twofold improvement in lesion detection on
3-T compared with 1.5-T MRI.
Because we could not control for the timing
of the 1.5-T and 3-T examinations, there might
have been differences in visualization of cer­
tain conditions that change over time, such
as tumor growth and conspicuity changes re­
lat­ed to interim developmental myelination
[17, 18]. We also did not blind individual
radiologists to viewing both sets of MR im­
ages at the same time. Although this factor
can introduce bias, it was unavoidable given
the nature of our study, which was direct
assessment of differences in image quality.
Identical protocol, acquisition times, and
coil type were not practically possible in this
retrospective study. We do not routinely use
dedicated surface coils or multichannel coils
for our 1.5-T MRI units in part because of
image degradation from inadequate signal-
to-noise ratio. Most of the patients in our
practice did not have specific localizing
information, rendering accurate surface coil
placement impossible. Phased-array surface
coils in the past have been limited by smaller
volumes of coverage and inherently hetero­
geneous signal intensity over the field of
view, precluding visualization of anatomic
detail of the brain parenchyma outside the
coil isocenter. Surface coils work well in the
evaluation of mesial temporal sclerosis, for
example, in which technologists can routinely
position the coil over the temporal lobes in
a reproducible manner. Finally, given the
high proportion of pediatric epilepsy in our
practice, a large number of patients undergo
sedation or anesthesia for MRI, making
coil changes during a whole-brain epilepsy
examination undesirable.
In the assessment of new imaging techno­
logy, an important consideration is whether
improvements in image quality have a
clinically beneficial effect [9]. Most patients
with focal epilepsy in this study had superior
lesion localization and diagnosis with 3-T
MRI. Most of the patients who underwent
surgery had substantial clinical improvement
or resolution of seizures.
In conclusion, MRI at 3 T was superior to
1.5-T MRI in the detection and accurate
 MRI of Epilepsy

characterization of structural brain abnor­
malities in a group of patients undergoing
whole-brain epilepsy evaluation at our
institution. Compared with 1.5-T MR images,
whole-brain 3-T MR images are of better
Downloaded from www.ajronline.org by 103.213.128.111 on 02/26/19 from IP address 103.213.128.111. Copyright ARRS. For personal use only; all rights reserved
5. Langfitt JT, Holloway RG, McDermott MP, et al.
Health care costs decline after successful epilepsy
surgery. Neurology 2007; 68:1290–1298
6. Urbach H, Hattingen J, von Oertzen J, et al. MR
imaging in the presurgical workup of patients
partial epilepsy: value of high-resolution volumetric
techniques. Am J Neuroradiol 1995; 16:339–343
13. Pattany PM. 3T MR imaging: the pros and cons.
Am J Neuroradiol 2004; 25:1455–1456
14. Frayne R, Goodyear BG, Dickhoff P, Lauzon

quality, do not require surface coils or
specific knowledge of lesion location, and
are not limited by technical artifacts. Imaging
at 3 T should be strongly considered in the
evaluation of patients with focal epilepsy and
previously equivocal or normal findings on
1.5-T MRI.
References
1. Von Oertzen J, Urbach H, Jungbluth S, et al. Stan­
dard magnetic resonance imaging is inadequate
for patients with refractory focal epilepsy. J Neu-
rol Neurosurg Psychiatry 2002; 73:643–647
2. Engel J Jr. Surgery for seizures. N Engl J Med
1996; 334:647–652
3. Wiebe S, Blume WT, Girvin JP, Eliasziw M. A ran­
domized, controlled trial of surgery for temporal-
lobe epilepsy. N Engl J Med 2001; 345: 311–318
4. Cascino GD. Improving quality of life with epi­
lepsy surgery: the seizure outcome is the key to
success. Neurology 2007; 68:1967–1968
with drug-resistant epilepsy. Am J Neuroradiol
2004; 25:919–926
7. Wyllie E, Lachhwani DK, Gupta A, et al. Suc­
cessful surgery for epilepsy due to early brain le­
sions despite generalized EEG findings. Neurolo-
gy 2007; 69:389–397
8. Urbach H. Imaging of the epilepsies. Eur Radiol
2005; 15:494–500
9. Vattipally VR, Bronen RA. MR imaging of epi­
lepsy: strategies for successful interpretation.
Neuroimaging Clin N Am 2004; 14:349–372
10. Knake S, Triantafyllou C, Wald LL, et al. 3T
phased array MRI improves the presurgical eval­
uation in focal epilepsies: a prospective study.
Neurology 2005; 65:1026–1031
11. Grant PE, Barkovich AJ, Wald LL, Dillon WP,
Laxer KD, Vigneron DB. High-resolution sur­
face-coil MR of cortical lesions in medically re­
fractory epilepsy: a prospective study. Am J Neu-
roradiol 1997; 18:291–301
12. Barkovich AJ, Rowley HA, Andermann F. MR in
ML, Sevick RJ. Magnetic resonance imaging at
3 Tesla: challenges and advantages in clinical
neurological imaging. Invest Radiol 2003; 38:
385–402
15. Schmitz BL, Aschoff AJ, Hoffmann MH, Gron G.
Advantages and pitfalls in 3T MR brain imaging:
a pictorial review. Am J Neuroradiol 2005;
26:2229–2237
16. DeLano MC, Fisher C. 3T MR imaging of the
brain. Magn Reson Imaging Clin N Am 2006; 14:
77–88
17. Eltze CM, Chong WK, Bhate S, Harding B, Nev­
ille BG, Cross JH. Taylor-type focal cortical dys­
plasia in infants: some MRI lesions almost disap­
pear with maturation of myelination. Epilepsia
2005; 46:1988–1992
18. Sankar R, Curran JG, Kevill JW, Rintahaka PJ,
Shewmon DA, Vinters HV. Microscopic cortical
dysplasia in infantile spasms: evolution of white
matter abnormalities. Am J Neuroradiol 1995;
16:1265–1272

F O R YO U R I N F O R M AT I O N
Unique customized medical search engine service from ARRS! ARRS GoldMiner™ is a keyword- and
concept-driven search engine that provides instant access to radiologic images published in peer-reviewed
journals. For more information, visit http://goldminer.arrs.org.

AJR:191, September 2008 895