Nonconvulsive Partial Status Epilepticus Mimicking Recurrent
Infarction Revealed by Diffusion-weighted and Arterial Spin
Labeling Perfusion Magnetic Resonance Images
Yuka Kanazawa, MD,* Takato Morioka, MD, PhD,† Shuji Arakawa, MD, PhD,*
Yoshihiko Furuta, MD,* Asako Nakanishi, MD,‡ and Takanari Kitazono, MD, PhDx

‘‘Non-convulsive’’ partial status epilepticus (SE) is an important pathologic condition that should be differentiated from cerebral infarction. Herein, we reported 2
patients who had partial SE associated with old infarction in the right parietal
lobe. Each patient had 2 episodes of left hemiparesis and hemisensory disturbance
without convulsion. On diffusion-weighted magnetic resonance images (DW-MRI),
a hyperintense lesion was noted in the cortex around the old infarction lesion, and
recurrent infarction was suspected. Although electroencephalography (EEG) failed
to reveal ictal discharges or interictal paroxysmal activities in 3 of 4 episodes, perfusion images with arterial spin labeling (ASL) clearly demonstrated ictal hyperperfusion in the area corresponding to the cortical hyperintense lesion on DW-MRI. After
appropriate anticonvulsant treatment based on the diagnosis of partial SE, clinical
symptoms were completely improved. These data stress the importance of cortical
hyperintensity on DW-MRI and ictal ASL hyperperfusion, even when SE cannot
be determined from EEG. Key Words: Stroke mimics—ictal hyperperfusion—
nonconvulsive status epilepticus—diffusion-weighted image—arterial spin
labeling.
Ó 2015 by National Stroke Association

Epilepsy is the most common disorder that can mimic
signs of stroke.1 Although ‘‘convulsive’’ status epilepticus
(SE) is easily recognized, inhibitory seizures such as
‘‘non-convulsive’’ prolonged partial epilepsy and partial
SE are difficult to differentiate from stroke.2 In acute
stroke, diffusion-weighted magnetic resonance images
From the *Department of Cerebrovascular Disease, Kyushu Rosai
Hospital, Kitakyushu; †Department of Neurosurgery Kyushu Rosai
Hospital, Kitakyushu; ‡Department of Radiology, Kyushu Rosai Hospital, Kitakyushu; and xDepartment of Medicine and Clinical Science,
Graduate School of Medical Sciences, Kyushu University, Fukuoka,
Japan.
Received July 28, 2014; revision received September 18, 2014;
accepted September 21, 2014.
The study received no funding or support.
Address correspondence to Yuka Kanazawa, MD, Department of
Cerebrovascular Disease, Kyushu Rosai Hospital, 1-1 Sonekitamachi,
Kokuraminami-ku, Kitakyushu, Fukuoka 800-0296, Japan. E-mail:
yuka8850@gmail.com.
1052-3057/$ - see front matter
Ó 2015 by National Stroke Association
http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.09.026

(MRI) (DW-MRI) and perfusion MRI (PI) are typically
used, although recent studies demonstrated that these
techniques provide information in the peri-ictal phase in
epilepsy patients.3-5 In SE, the epileptogenic cortex is in
an electrophysiologically extreme state, and the activated
cortex exhibits increased glucose and oxygen usage,
thereby causing compensatory regional hyperperfusion. Arterial spin labeling (ASL) is a noninvasive and
repeatable PI technique, which uses magnetically labeled
blood water as an endogenous tracer.6 Recent reports
have described the appearance of ‘‘ictal hyperperfusion’’
with ASL.6-10 When the hyperperfusion is no longer
sufficient to supply the hyperactive cortical area,
pathophysiologic changes leading to cytotoxic edema in
epileptic cortical neurons can occur, which appear as an
abnormal high signal in the cortical lamina (cortical
hyperintensity) on DW-MRI.3,7,11-15 These MRI findings
of low apparent diffusion coefficient and high signals
on DW-MRI in SE resemble those of acute ischemic
stroke, indicating changes attributable to both cytotoxic
and vasogenic edema.4,7 Such ictal changes on ASL and

Journal of Stroke and Cerebrovascular Diseases, Vol. 24, No. 4 (April), 2015: pp 731-738

731

Y. KANAZAWA ET AL.

732
5,7

DW-MRI were reversible in most cases. In the present
study, using serial MRI including DW-MRI and ASL
and electroencephalography (EEG) recording, we examined 2 patients with nonconvulsive partial SE clinically resembling recurrent infarction around the old
infarction.

Methods
Magnetic Resonance Image
Brain MRI with routine protocols and PI were performed using a 3-T magnetic resonance unit (Signa HDxt
3.0T, version 23; GE Healthcare, Milwaukee, WI). Routine
protocols included axial diffusion-weighted echo planar
sequences (b value 5 1500 seconds/mm2; repetition time
(TR)/echo time (TE), 6000/min), T1 fluid-attenuated
inversion recovery (T1-FLAIR) sequences (TR/TE/TI,
2050/16.1/AUTO), and T2-weighted fast spin-echo
sequences (TR/TE, 4400/100) and T2-FLAIR sequence
(TR/TE/TI, 9000/140/AUTO).
ASL was prepared using 3-dimensional, spiral, fast
spin-echo sequence with background suppression for
perfusion imaging covering the entire brain. A pulsed
continuous scheme was employed. Other acquisition parameters were as follows: 4 arms with 1004 points in each
spiral arm, phase encoding in the z direction 5 32, section
thickness 5 4 mm, TR 5 4728(AUTO) seconds, postlabel
wait 5 1.525 seconds, and number of excitations 5 3.

Electroencephalography
Routine EEG recordings were obtained from an
18-channel digital EEG machine (Neurofax; NihonKohden, Tokyo, Japan) with electrode placement according to the International EEG 10-20 system. The EEG
recordings were performed for at least 30 minutes for
each patient in resting conditions.

Results

(edaravone) were administrated intravenously. The left
hemiparesis and hemisensory disturbance were gradually improved, and DW-MRI on day 4 demonstrated
dramatic disappearance of cortical hyperintensity in the
right parietal lobe (Fig 1, C). However, ASL clearly
showed hyperperfusion in the corresponding area to
that of the cortical hyperintensity on DW-MRI of day 1
(Fig 1, D). Diagnosis of symptomatic partial epilepsy was made and intravenous phenytoin followed
by oral carbamazepine was administered. On day 5,
she completely recovered from left hemiparesis and sensory disturbance, and EEG demonstrated intermittent
focal slow wave on the right centroparieto-occipital
region (Fig 2, A).
Unfortunately, the patient chose to discontinue carbamazepine treatment when she was transferred to the
orthopedic department. Ten months later, she again
developed left hemiparesis and hemisensory disturbance
(day 1’, day 1 of the second episode). On DW-MRIs, the
cortical hyperintense lesion was again noted around the
old infarction in the right parietal lobe (Fig 1, E). ASL
clearly showed hyperperfusion in the corresponding
area to that of the cortical hyperintensity on DW-MRI
(Fig 1, F). These DW-MRI and ASL findings were quite
similar to those at the first episode. Thus, intravenous fosphenytoin followed by oral carbamazepine was administered. Although EEG failed to reveal ictal discharges,
frequent interictal paroxysmal activities were noted on
the right centroparietal region (Fig 1, B; P4 and C4), which
was identical to the area of cortical hyperintensity on
DW-MRI and hyperperfusion on ASL. On day 2’, she
completely recovered from left hemiparesis and sensory
disturbance, and EEG of day 6’ demonstrated occasional
low-amplitude paroxysmal activities in the right centroparietal region (Fig 2, C). With monotherapy of carbamazepine, she was free from epilepsy during the 2 years
following her second episode. The cortical hyperintensity
on DW-MRI was transient, and subsequent T2 prolonged
lesion was not demonstrated on the follow-up images.

Case 1
A 74-year-old female developed left hemiparesis and
was admitted to us. She was alert on arrival (day 1).
She had left hemiparesis and sensory disturbance in
the left extremities, although convulsive seizures were
not apparent. Emergency MRI with FLAIR revealed multiple old infarctions in the right parietal lobe and deep
white matter on the both side (Fig 1, A). The cortex of
the old infarction area in the right parietal lobe exhibited
cortical laminar necrosis. On DW-MRIs, a hyperintense
lesion was noted in the cortex around the old infarction
site in the right parietal lobe (Fig 1, B). The precentral
gyrus was not involved in this cortical hyperintensity.
The tentative diagnosis was recurrent infarction around
the old right parietal infarction, and an antithrombin
agent (argatroban hydrate) and a free radical scavenger

Case 2
A 74-year-old female with a past history of infarction in
the right parietal lobe developed left hemiparesis and
sensory disturbance (day 1). She was alert on admission.
She had left hemiparesis and sensory disturbance in the
left extremities and dressing apraxia, although convulsive
seizures were not apparent. FLAIR revealed old infarctions in the right parietal lobe (Fig 3, A). Perilesional gliosis was also noted. On DW-MRI, a gyriform cortical
hyperintense lesion was noted around the old infarction
in the right parietal lobe (Fig 3, B). The precentral gyrus
was not involved. Her abnormal neurological findings
were gradually improved. Although epilepsy was highly
suspected, antiepileptic drug was not administered as
EEG failed to reveal paroxysmal discharges (Fig 4, A).

STATUS EPILEPTICUS AS STROKE MIMICS

733

Figure 1. Case 1. (A) Magnetic resonance images (MRI) with fluid-attenuated inversion recovery sequence (FLAIR) on admission (day 1) demonstrated an old
infarction as cortical laminar necrosis in the right parietal lobe (solid arrows). Multiple cerebral infarctions were also noted in the bilateral white matter. (B)
Diffusion-weighted MRI (DW-MRI) demonstrated gyriform cortical hyperintensity around the old infarction (dotted arrows). The precentral gyrus was not
involved with cortical hyperintensity. (C) DW-MRI on day 4 indicated that the cortical hyperintensity had almost disappeared. (D) Perfusion images with arterial
spin labeling (ASL) on day 4 showed hyperperfusion in the right parietal lobe that corresponds to the area of cortical hyperintensity on DW-MRI of day 1. (E) On
DW-MRIs at the onset (day 1’) of the second episode (10 months after the first episode), cortical hyperintensity lesion was noted around the old infarction in the
right parietal lobe. (F) ASL clearly showed hyperperfusion in the corresponding area to that of cortical hyperintensity on DW-MRI.

734

Y. KANAZAWA ET AL.

Figure 2. Case 1. (A) Electroencephalogram (EEG) on day 5 showed intermittent focal slow waves on the right centroparieto-occipital region (black lines). (B)
EEG on day 1’ failed to reveal ictal discharges, whereas frequent interictal paroxysmal activities were noted on the right centroparietal region (P4 and C4 of International EEG 10-20 System), which is identical to the area of cortical hyperintensity on DW-MRI and hyperperfusion on ASL. (C) EEG on day 6’ revealed
occasional small-amplitude paroxysmal activities on the right centroparietal region (arrows).

STATUS EPILEPTICUS AS STROKE MIMICS

735

Figure 3. Case 2. (A) Fluid-attenuated inversion recovery sequence of day 2 revealed old infarctions in the right parietal lobe. Perilesional gliosis was also noted.
(B) On DW-MRI, a gyriform cortical hyperintense lesion was noted around the old infarction in the right parietal lobe. The precentral gyrus was not involved with
the cortical hyperintensity. (C) DW-MRI on day 6 demonstrated a dramatic disappearance of cortical hyperintensity in the right parietal lobe. (D) On DW-MRI at
the onset (Day 1’) of second episode (4 months after the first episode), a cortical hyperintense lesion was noted around the old infarction in the right parietal lobe.
(E) On DW-MRI of day 2’ the cortical hyperintensity had disappeared, whereas ASL still depicted hyperperfusion in the area corresponding to that of cortical
hyperintensity on DW-MRI of day 1’.

736

Y. KANAZAWA ET AL.

Figure 4. Case 2. (A) Electroencephalogram (EEG) of day 3 failed to reveal paroxysmal discharges. Predominantly continuous slow wave activities were noted
in the right parietal region (black line). (B) EEG of day 2’ showed focal slow waves on the right centroparietal region (black lines).

Predominant continuous slow wave activities were noted
on the right parietal region. DW-MRI on day 6 demonstrated dramatic disappearance of cortical hyperintensity
in the right parietal lobe (Fig 3, C).
Four months later, she developed left hemiparesis and
sensory disturbance again. On DW-MRI on day 1’ (day 1
of the second episode), cortical hyperintensity lesion was
noted around the old infarction in the right parietal lobe

(Fig 3, D). With administration of carbamazepine, her
abnormal neurologic findings were improved. EEG on
day 2’ failed to reveal paroxysmal discharges and showed
focal slow waves on the right centroparietal region (Fig 4,
B). On DW-MRI of Day 2’, the cortical hyperintensity had
disappeared, while ASL still indicated hyperperfusion in
the corresponding area to that of cortical hyperintensity
on DW-MRI of day 1’ (Fig 3, F). With monotherapy of

STATUS EPILEPTICUS AS STROKE MIMICS

carbamazepine, she was free from epilepsy during the
2 years from her second episode. Subsequent T2 prolonged lesion was not noted on the follow-up images.

Discussion
Although ‘‘convulsive’’ SE is easily recognized, in
the diagnosis of partial SE without discrete convulsion
or such as our cases, capture of the ictal discharge on
EEG is considered the gold standard.12,16,17 However,
it is practically difficult to record the ictal discharges
without continuous EEG monitoring facilities. In
most hospital, routine EEG examination is unavailable
outside working hours or on weekends, and the timing
of EEG recording is often delayed.16,17 In the present
study, we could not obtained ictal discharges in all 4
episodes of our 2 cases. A further problem in the EEG
diagnosis of epilepsy is the sensitivity of post- or interictal EEG. During interictal state, paroxysmal activities
are not always recorded. For adults presenting with a first
seizure, a routine EEG revealed paroxysmal activities in
only approximately 23% of patients.18 In our study, interictal paroxysmal discharges were recorded only on day 1’
in the second episode of case 1. Furthermore, a marked
decrease in amplitude of these paroxysmal activities
was noted on day 6’, and these subtle paroxysms may
have been overlooked. In other 3 episodes of our 2 cases,
localized slow wave activities on the area corresponding
to the old infarction were observed (only indicative of
focal dysfunction) as the timing of EEG recordings was
obviously delayed.
Theoretically, ictal hyperperfusion is often followed by
postictal hypoperfusion, whereas temporal evolution of
regional ictal hyperperfusion in the postictal stage has
not been clearly defined. On ictal/postictal perfusion
single photon emission tomography, Lee et al19 demonstrated that postictal perfusion abnormalities did not return to the interictal phase at 6 hours after the epilepsy.
A previous ASL study also reported that postictal regional
hyperperfusion was observed immediately to 5 hours
after a single seizure.6,8,9 In the first episode of our case
1 and the second episode of case 2, postictal ASL
hyperperfusion was still observed on day 4 and day 2’,
respectively, when the cortical hyperintensity on DWMRI had almost disappeared. Oishi et al7 reported that
postictal ASL hyperperfusion was observed in the medial
to lateral cortex of the right occipital lobe on the day of
partial SE, when the hyperintensity on DW-MRI was
observed in a limited area of the medial occipital lobe.
These findings suggest that ictal ASL hyperperfusion
persists postictally longer than that for DW-MRI hyperintensity, although this likely depends on the magnitude
and duration of the epileptic activities.4,5 Thus, the
combined use of peri-ictal ASL and DW-MRI4,5,7,20 may
overcome the limitations of peri-ictal EEG recording
in the diagnosis of nonconvulsive partial SE.

737

Cortical hyperintensity on DW-MRI should be differentiated from (pseudo) cortical laminar necrosis (CLN)
associated with cerebral infarction. Incomplete infarction
results in selective neuronal loss in the cortex with the
presence of viable cells, glial proliferation, and deposition
of paramagnetic substances. In acute infarction, CLN is
demonstrated as a gyriform cortical high signal intensity
area on DW-MRI,21 which is quite similar to cortical hyperintensity associated with SE. CLN is an irreversible
change and the chronological changes have been well
demonstrated on various sequences of serial MRI.22
In our case 1, the cortex of the old infarction area in the
right parietal lobe was demonstrated as CLN on FLAIR
image. By contrast, cortical hyperintensity associated
SE is transient.4,5,7,20 Furthermore, the area of cortical
hyperintensity on DW-MRI should correspond to the hyperperfused area,4,5,20 because the ictal hyperperfusion is
a secondary change to meet the increased glucose and
oxygen demand of the activated cortex. In both our
cases, ASL clearly showed hyperperfusion in the area
corresponding to that of cortical hyperintensity on
DW-MRI at the ictus. Nonconvulsive partial SE can be
complicated with stroke both in the acute and chronic
stages.12,16,17 In the present cases, cortical hyperintensity
on DW-MRI was transient and subsequent T2 prolonged
lesion was not demonstrated on the follow-up images.
Therefore, the recurrence of cerebral infarction can be
ruled out.
Another advantage of the combined use of ASL
and DW-MRI is ability to use the localization data to
document the pathophysiologic mechanism of partial
epilepsy in each patient.7,10,20 In our 2 patients, the ASL
hyperperfusion and cortical hyperintensity on DW-MRI
were observed around the old infarction. This finding
clearly depicted the topographic relationship between
the epileptogenic lesion (the old infarction) and the activated cortex (right parietal lobe around the old infarction). With this method, it is impossible to localize the
epileptogenic focus or area, but the cortical area involved
by spread of the epileptic activities in prolonged epilepsy
can be clearly localized.5 Our finding that the precentral
gyrus was not involved with the epileptic activities may
explain why our 2 patients developed ‘‘non-convulsive’’
SE instead of ‘‘convulsive’’ SE.
In conclusion, combined use of DW-MRI and ASL can
provide information complementary to high-resolution
structural MRI in the peri-ictal phase of nonconvulsive
partial SE that may overcome the limitations of periictal EEG recording and help differentiate this disorder
from acute ischemic stroke.

References
1. Winkler DT, Fluri F, Fuhr P, et al. Thrombolysis in stroke
mimics: frequency, clinical characteristics, and outcome.
Stroke 2009;40:1522-1525.

738
2. De Reuck J, Van Maele G. Transient ischemic attacks and
inhibitory seizures in elderly patients. Eur Neurol 2009;
62:344-348.
3. Di Bonaventura C, Bonini F, Fattouch J, et al. Diffusionweighted magnetic resonance imaging in patients with
partial status epilepticus. Epilepsia 2009;50:45-52.
4. Flacke S, Wullner U, Keller E, et al. Reversible changes in
echo planar perfusion- and diffusion-weighted MRI in
status epilepticus. Neuroradiology 2000;42:92-95.
5. Szabo K, Poepel A, Pohlmann-Eden B, et al. Diffusion
weighted and perfusion MRI demonstrates parenchymal
changes in complex partial status epilepticus. Brain 2005;
128:1369-1376.
6. Pizzini FB, Farace P, Manganotti P, et al. Cerebral perfusion alterations in epileptic patients during peri-ictal
and post-ictal phase: PASL vs DSC-MRI. Magn Reson
Imaging 2013;31:1001-1005.
7. Oishi M, Ishida G, Morii K, et al. Ictal focal hyperperfusion demonstrated by arterial spin-labeling perfusion
MRI in partial epilepsy status. Neuroradiology 2012;
54:653-656.
8. Pizzini F, Farace P, Zanoni T, et al. Pulsed-arterial-spinlabeling perfusion 3T MRI following single seizure: a first
case report study. Epilepsy Res 2008;81:225-227.
9. Pollock JM, Deibler AR, West TG, et al. Arterial spinlabeled magnetic resonance imaging in hyperperfused
seizure focus: a case report. J Cmput Assist Tomogr
2008;32:291-292.
10. Toledo M, Munuera J, Salas-Puig X, et al. Localisation
value of ictal arterial spin-labelled sequences in partial
seizures. Epileptic Disord 2011;13:336-339.
11. Diehl B, Najm I, Ruggieri P, et al. Periictal diffusionweighted imaging in a case of lesional epilepsy. Epilepsia
1999;40:1667-1671.
12. Hamamura T, Morioka T, Sayama T, et al. Cerebral hyperperfusion syndrome associated with non-convulsive
status epilepticus following superficial temporal arterymiddle cerebral artery anastomosis. Case report. Neurol
Med Chir (Tokyo) 2010;50:1099-1104.

Y. KANAZAWA ET AL.
13. Kim J-A, Chung JI, Yoon PH, et al. Transient MR
signal changes in patients with generalized tonicoclonic seizure or status epilepticus: periictal diffusionweighted imaging. AJNR Am J Neuroradiol 2001;22:
1149-1160.
14. Lansberg MG, O’Brien MW, Norbash AM, et al. MRI
abnormalities associated with partial status epilepticus.
Neurology 1999;52:1021-1027.
15. Wieshmann UC, Symms MR, Shorvon SD. Diffusion
changes in status epilepticus. Lancet 1997;350:493-494.
16. Morioka T, Sayama T, Nukae N, et al. Nonconvulsive status epilepticus during perioperative period of cerebrovascular surgery. Neurol Med Chir (Tokyo) 2011;51:
171-179.
17. Morioka T, Sayama T, Shimogawa T, et al. Electroencephalographic evaluation of cerebral hyperperfusion syndrome following superficial temporal artery–middle
cerebral artery anastomosis. Neurol Med Chir (Tokyo)
2013;53:388-395.
18. Krumholz A, Wiebe S, Gronseth G, et al. Practical parameter: evaluating an apparent unprovoked first seizure
in adults (an evidence-based review): report of the
Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society.
Neurology 2007;69:1996-2007.
19. Lee DS, Lee SK, Kim SK, et al. Late postictal residual
perfusion abnormality in epileptogenic zone found on
6-hour postictal SPECT. Neurology 2000;55:835-841.
20. Calistri V, Caramia F, Bianco F, et al. Visualization of
evolving status epilepticus with diffusion and perfusion MR imaging. AJNR Am J Neuroradiol 2003;24:
671-673.
21. Lovblad KO, Wetzel SG, Somon T, et al. Diffusionweighted MRI in cortical ischaemia. Neuroradiology
2004;46:175-182.
22. Kesavadas C, Santhosh K, Thomas B, et al. Signal changes
in cortical laminar necrosis. Evidence from susceptibilityweighted magnetic resonance imaging. Neuroradiology
2009;51:293-298.