RESEARCH ARTICLE

Significance of Very-High-Frequency Oscillations (Over 1,000Hz) in Epilepsy Naotaka Usui, MD, Kiyohito Terada, MD, Koichi Baba, MD, Kazumi Matsuda, MD, Keiko Usui, MD, Takayasu Tottori, MD, Tadahiro Mihara, MD, and Yushi Inoue, MD Objective: We previously reported ictal very-high-frequency oscillations (VHFO) of 1,000 to 2,500Hz recorded by subdural macroelectrodes using a 10-kHz sampling rate. The purpose of this study was to clarify the clinical significance of ictal VHFO in neocortical epilepsy. Methods: This study included 13 patients with neocortical epilepsy who underwent subdural electrode implantation and had at least 1 seizure recorded at a 10-kHz sampling rate and were followed for more than 2 years postoperatively. Extent of resection was determined considering the seizure onset zone (SOZ) and irritative zone, structural lesion, and functional areas. Areas showing VHFO and those with HFO were not taken into consideration. The presence or absence of VHFO (>1,000Hz), HFO (200–1,000Hz) and SOZ, and completeness of resection of these areas were compared with postoperative seizure outcome. Results: Seven patients had favorable (Engel class Ia) and 6 had unfavorable outcomes (other classes). VHFO was recorded in 6 of 7 patients with a favorable outcome. On the contrary, VHFO was recorded in only 1 of 6 patients with unfavorable outcome. The presence of VHFO was significantly associated with favorable outcome. VHFO was recorded on a limited number of electrodes, and VHFO-generating areas were resected completely, whereas HFO-generating areas and/or SOZ were not always resected completely in both favorable and unfavorable outcome groups. Interpretation: The presence of ictal VHFO may be predictive of favorable outcome. Ictal VHFO may be a more specific marker than ictal HFO or SOZ for identifying the core of epileptogenic zone. ANN NEUROL 2015;78:295–302

I

ncreasing attention is being paid to high-frequency activities recorded on intracranial electroencephalogram (EEG). High-frequency oscillations (HFOs), such as ripples (100–250Hz) and fast ripples (250–500Hz), have been mainly investigated, and the usefulness of HFOs for deciding the extent of surgical resection has been suggested.1–3 Akiyama et al reported that more-complete resection of the regions with high-rate fast ripples significantly correlated with better seizure outcome.3 Recently, we reported on the presence of very-highfrequency oscillations (VHFOs) of over 1,000Hz (1,000– 2,500Hz) recorded by subdural electrodes at a very high sampling rate of 10kHz in patients with neocortical epilepsy. Compared to HFOs, VHFOs have much higher frequency, lower amplitude, more restricted distribution, and different timing of onset. VHFOs and HFOs may have different pathophysiological mechanisms.4

To further clarify the clinical significance of VHFOs, especially for determining the optimal extent of resection in epilepsy surgery and for predicting postoperative seizure outcome, we conducted a retrospective study in 13 patients with intractable neocortical epilepsy followed for more than 2 years after surgical resection. We compared postoperative seizure outcome with the presence or absence of ictal VHFOs, ictal HFOs, and seizure onset zone (SOZ) defined by conventional EEG recordings, and completeness of resection of the above areas.

Patients and Methods Thirteen patients (7 males and 6 females) with intractable neocortical epilepsy, who were monitored electroencephalographically at a sampling rate of 10kHz before resection surgery and were followed for more than 2 years postoperatively, were included in this study. Clinical characteristics of the patients are

View this article online at wileyonlinelibrary.com. DOI: 10.1002/ana.24440 Received Jan 7, 2015, and in revised form Apr 22, 2015. Accepted for publication May 10, 2015. Address correspondence to Dr Naotaka Usui, National Epilepsy Center, Shizuoka Institute of Epilepsy and Neurological Disorders, Urushiyama 886, Aoi-ku, Shizuoka, 420–8688 Japan. E-mail: [email protected] From the National Epilepsy Center, Shizuoka Institute of Epilepsy and Neurological Disorders, Shizuoka, Japan.

C 2015 American Neurological Association V 295

ANNALS

of Neurology

shown in Table 1. Age at seizure onset ranged from 0 to 20 (mean, 6) years and age at surgery from 6 to 57 (mean, 24) years. All patients underwent intracranial EEG monitoring as a part of presurgical evaluation. Subdural electrodes (2.3 mm contact, effective area 4.15 mm2, 10 mm spacing, platinum/ iridium alloy; Ad-tech Medical Instrument Corporation, Racine, WI) were implanted over the cortical areas depending on the findings of noninvasive presurgical evaluation. Reference electrodes were placed on the surface of the skull, with the contacts of the electrodes facing away from the skull to avoid referential activation. EEG signals were digitally recorded by EEG1000 (Nihon-Kohden Corporation, Tokyo, Japan) at a sampling rate of 200Hz and high-pass filter of 0.016Hz for routine clinical evaluation. The Butterworth type of filter was used for the EEG machine. After completing the routine EEG recording, EEG recording using a higher sampling rate of 10kHz was conducted. The high-pass filter was set at 0.016Hz. Owing to the limitation of the EEG machine, only 16 channels could be monitored simultaneously. Sixteen channels were selected based on the conventional EEG findings, basically considering the SOZ and/or irritative zone, and structural lesion on magnetic resonance imaging (MRI), if present. To visualize highfrequency activities, horizontal (time) and vertical (amplitude) axes of the EEG display were expanded, and EEG was digitally high-pass filtered at 160Hz and low-pass filtered at 3kHz. Peaks of high-frequency activities were visually identified on a cathode ray tube screen, and frequency, amplitude, and duration of these activities were measured by cursors with computer assistance. To exclude the possibility of a “false” high-frequency component produced by filter, EEG data high-pass filtered at 160Hz and those at 53Hz were compared. Consequently, frequencies of VHFOs observed at both filter settings were the same, and the waveforms were considered as true EEG activities. Ictal EEGs were analyzed visually by two clinical epileptologists (N.U. and K.T.). Both observers jointly reviewed the data and established a consensus. High-frequency activities faster than 200Hz and slower than 1,000Hz were defined as HFO (fast ripples) and those faster than 1,000Hz as VHFO. Activities slower than 200Hz (ripples) were not analyzed in this study. As reported previously, intermittent VHFOs appeared inter- and preictally, and sustained VHFO appeared after ictal onset.4 The location of VHFO in relation to the structural lesion was also examined. SOZ was defined by conventional EEG recordings as the contact that showed the first ictal activity, preceding or concomitant to clinical onset of seizures. Widespread or diffuse voltage attenuation was excluded from ictal EEG changes. Irritative zone was defined as cortical area that was generating interictal spikes. Extent of resection was determined considering SOZ and irritative zones determined by conventional EEG recordings, structural lesions, and functional areas. Areas showing VHFOs and HFOs were not taken into consideration for surgical decision making. Postoperative follow-up ranged from 2 to 4 years. Seizure outcome was defined by Engel’s criteria. Favorable

296

outcome was defined as class Ia and unfavorable outcome as other classes (classes Ib, c, d, II, III, and IV). The presence or absence VHFO (>1,000Hz), HFO (200–1,000Hz) and SOZ, as well as completeness of resection of these areas, were compared with postoperative seizure outcome, using Fisher’s exact test. Mann–Whitney U test was employed in comparing resection rates. The statistically significant level was p < 0.05.

Results Table 2 shows the number of seizures recorded with a 10-kHz sampling rate, total number of electrodes implanted, number of electrodes showing ictal VHFO or ictal HFO, number of SOZ electrodes, percent resection of VHFO-generating area (residual VHFO 5 No. of electrodes with VHFO outside resection margin/Total No. of electrodes with VHFO 3 100(%)), that of HFOgenerating area (residual HFO 5 No. of electrodes with HFO outside resection margin/Total No. of electrodes with HFO 3 100(%)), that of SOZ (residual SOZ 5 No. of electrodes within SOZ outside resection margin/Total No. of electrodes within SOZ 3 100(%)), completeness of MRI lesion resection, pathology of resected specimens, follow-up period, and postoperative seizure outcome. Favorable seizure outcome was obtained in 7 patients, and unfavorable in the remaining 6. VHFOs were detected in 7 patients. VHFOs were recorded from 1 to 4 electrodes in each patient. The representative VHFOs of Patients 2 and 8 recorded on the only channel showing VHFO are shown in Figures 1 and 2. Preictally, VHFO appeared intermittently, interrupted by spikes. Sustained VHFO without spikes appeared from around the start of seizures. Waveforms of VHFOs are not typical sinusoidal waves. Fluctuations in frequency and amplitude for individual peaks appear in VHFO waveforms. Frequencies of VHFOs were 1,000 to 1,500Hz in 5 patients, 1,000 to 2,000Hz in 1, and 1,000 to 2,500Hz in another. Amplitudes of VHFO ranged from 3.5 to 29.4lV and durations from 2 to 226msec. VHFOs were always recorded within the SOZ and originated just above hyperintense regions on MRI. Areas showing VHFO were resected completely in all 7 patients. VHFOs were detected in 6 of 7 patients with favorable outcomes, but in only 1 of 6 with unfavorable outcomes. Presence of VHFO was significantly associated with favorable seizure outcome. Ictal HFO were detected in 10 patients. They were detected in 6 of 7 patients with favorable outcome and in 4 of 6 with unfavorable outcome. Frequency of HFO ranged from 200 to 900Hz, amplitudes from 10.3 to 279.4lV, and durations from 6 to 59msec. Presence of HFO was not associated with seizure outcome. HFOs Volume 78, No. 2

August 2015

FLE

FLE

FLE

PLE

PCE

FLE

FLE

PLE

FLE

FLE

PRLE

PE

PCE

1

2

3

4

5

6

7

8

9

10

11

12

13

7

2

14

0

0

4

20

5

4

10

5

4

3

Age at Onset

12

44

18

6

22

14

57

31

19

25

14

27

19

Age at Surgery

Generalized jerking

Somatosensory

Somatosensory!generalized jerking

Aura!tonic posturing

Tonic posturing

Tonic posturing

Autonomic

Aura!gestural automatism

Lt arm clonic!generalized clonic

Tonic posturing

Tonic facial contraction

Gestural automatism

Gestural automatism

Seizure

Lt pT, Rt P-O, and Rt F-C spikes

Rt F-T and pT-O spikes

None

Lt F-T spikes

Lt central spikes

Rt parietal sharp waves

Normal

Lt frontal spikes

Rt occipital and frontal spikes

Rt parietal spikes

Lt frontal sharp waves

Rt F-T spikes

Normal

Interictal Scalp EEG

Nonlateralizing

Nonlateralizing

Rt central

Lt frontal

Nonlateralizing

Rt hemisphere

Nonlateralizing

Lt frontal

Rt posterior

Rt parietal

Lt frontal

Rt F-T

Rt hemisphere

Ictal Scalp EEG

Blt P-O

Rt pariretal-insula

Rt paracentral lobule

Normal

Lt posterior frontal

Rt medial parietal

Lt medial frontal

Normal

Rt P-O

Rt supramarginal

Lt frontal operculum

Rt frontal operculum

Rt basal frontal

MRI

FLE 5 frontal lobe epilepsy; PCE 5 posterior cortex epilepsy; PRLE 5 perirolandic epilepsy; PE 5 partial epilepsy; Rt 5 right; Lt 5 left; Blt 5 bilateral; T-O 5 temporooccipital; PO 5 parietooccipital; F-C 5 frontocentral; F-T 5 frontotemporal.

Diagnosis

Patient

TABLE 1. Clinical Characteristics of the Patients

Usui et al: Significance of VHFO in epilepsy

297

298

4

3

2

3

1

10

2

2

2

2

1

2

3

1

2

3

4

5

6

7

8

9

10

11

12

13

56

56

48

74

38

44

84

110

66

56

94

84

82

Total No. of Electrodes

0

0

0

0

0

3

1

4

2

0

4

1

2

No. of VHFO Electrodes

NA

NA

NA

NA

NA

0

0

0

0

NA

0

0

0

Residual VHFO (%)

6

0

4

10

0

10

2

6

14

7

6

9

0

No. of HFO Electrodes

0

NA

100

70

NA

40

50

0

29

57

33

11

NA

Residual HFO (%)

11

Not identified

Not identified

Not identified

4

8

1

6

10

16

6

5

4

No. of SOZ Electrodes

0

NA

NA

NA

0

25

0

0

60

69

0

0

0

Residual SOZ (%)

Incomplete

Incomplete

Complete

NA

Incomplete

Complete

Complete

NA

Complete

Complete

Complete

Complete

Complete

MRI Lesion Resection

2

3

3

3

4

3

2

2

3

3

3

4

4

Follow-up (Year)

IVa

IIIa

IIc

IVb

IIIa

Ib

Ia

Ia

Ia

Ia

Ia

Ia

Ia

Seizure Outcome

Ulegyria

FCD IIb

Glioma

NS

FCD IIb

FCD IIa

FCD IIb

FCD IIb

FCD IIb

FCD IIa

FCD IIb

FCD IIb

FCD IIa

Pathology

VHFO 5 very high frequency oscillations; HFO 5 high-frequency oscillations; SOZ 5 seizure onset zone; NA 5 not available; FCD 5 focal cortical dysplasia; NS 5 nonspecific.

No. of Seizures

Pt

TABLE 2. Results of VHFO, HFO, and SOZ From EEG Recordings and Outcome of Surgery

ANNALS

of Neurology

Volume 78, No. 2

Usui et al: Significance of VHFO in epilepsy

FIGURE 1: Waveform of very-high-frequency oscillations (VHFO) in Patient 2. Ictal electroencephalogram (EEG) recorded at a sampling rate of 10kHz. (A) Ictal EEG using conventional filter settings (low-pass filter of 120Hz and high-pass filter of 1.6Hz). The single channel that shows VHFO is shown. Conventional ictal EEG shows preictal spiking followed by low-voltage fast activity. The EEG at the points marked B, C, D, and E using VHFO filter settings are shown in separate panels (B) to (E). (B) to (D) Intermittent VHFO (underbars) observed using a low-pass filter of 3kHz and a high-pass filter of 160Hz. VHFO of 1,000 to 2500Hz are detected. They appeared before and around seizure onset. (E) Sustained VHFO detected after electrodecremental pattern. They superimpose on the slower rhythmic activity. Note that the interpeak durations and amplitudes of VHFO waveforms are variable.

were recorded from 4 to 14 electrodes in each patient. In patients with HFO, percent resection of HFO-generating areas was 43 to 100 (mean, 60) in 6 patients with favorable outcome and 0 to 100 (mean, 32) in 4 with unfavorable outcome. Percent resection did not differ significantly between patients with favorable outcome and those with unfavorable outcome. SOZ was not identified in 3 of 6 patients with unfavorable outcome. In these 3 patients, no EEG changes other than widespread or diffuse voltage attenuation were detected preceding or concomitant to clinical seizure onset. Identification of SOZ was not associated with seizure outcome. In 10 patients whose SOZ was identified, percent resection of the SOZ was 31 to 100 (mean, 82) in 7 patients with favorable outcome and 75 to 100 (mean, 46) in 3 with unfavorable outcome. Percent resection of SOZ was not significantly different between patients with favorable outcome and those with unfavorable outcome. MRI lesion resection was complete in 8 patients and incomplete in 3. No patients with incomplete lesion resection achieved seizure freedom. Histopathology of resected specimens was focal cortical dysplasia (FCD) type II in 10 of 13 patients. August 2015

Histopathology revealed FCD type II in all 7 patients with VHFO, type IIa in 2, and type IIb in 5.

Discussion Several recent articles reported that removal of HFOgenerating tissues correlated with good surgical outcome.1–3 However, the clinically significant frequency range remains undefined. HFOs are not associated with specific underlying pathology.5 We previously recorded low-amplitude VHFO (1,000–2,500Hz) by routinely using subdural electrodes at a 10-kHz sampling rate in 4 of 5 neocortical epilepsy patients.4 In the present study, we investigated the clinical significance of ictal VHFO, HFO, and SOZ in a larger series for a longer postoperative follow-up period, focusing on the relationship with postoperative seizure outcome. There are several points that differentiate VHFO from a false high-frequency component produced by the filter. Ringing artifact, which is caused by Gibbs phenomenon, will show different frequencies by different filter settings. On the contrary, VHFOs have the same frequencies by different filter settings, and it is considered that they are not the false high-frequency component produced by a filter. The second is that ringing 299

ANNALS

of Neurology

FIGURE 2: Waveform of very-high-frequency oscillations (VHFO) in Patient 8. Ictal electroencephalogram EEG recorded at a sampling rate of 10kHz. (A) Ictal EEG using conventional filter settings (low-pass filter of 120Hz and high-pass filter of 1.6Hz). The single channel that shows VHFO is shown. Conventional ictal EEG shows preictal spiking followed by low-voltage fast activity. The EEG at the points marked B, C, D, and E using VHFO filter settings are shown in separate panels (B) to (E). (B) to (D) Intermittent VHFO (underbars) observed using a low-pass filter of 3kHz and high-pass filter of 160Hz. VHFO of 1,000 to 2,000Hz are detected. They appeared before and around seizure onset. (E) Sustained VHFO detected after electrodecremental pattern. They superimpose on the slower rhythmic activity. Note that the interpeak durations and amplitudes of VHFO waveforms are variable.

artifact appears as a damped sinusoidal wave, whereas the waveform of VHFO is not a typical sinusoidal wave. VHFO waveforms have fluctuations in terms of frequency and amplitude. The third is the timing of appearance. Ringing artifact appears before or after the initial rising point of spikes, whereas VHFOs appear only before the initial rising point of spikes. Differentiating VHFO from electromyography (EMG) activity is another important issue. It is known that EMG activity can penetrate the bone defect, and high-frequency EMG activity can be recorded by localized intracranial electrodes. Otsubo et al recorded intracranial EEG using a sampling rate of 1kHz and underwent multiple-band frequency analysis (MBFA) of muscle impulses during chewing. MBFA showed scattered HFO without specific frequency band lasting 300 to 400ms.6 Differently from high-frequency EMG activity, intermittent VHFOs appear with spikes, and the duration of VHFO is very short (2 to 226ms). VHFOs can be differentiated from EMG activity by these points. Ictal VHFOs were recorded in 6 of 7 patients with favorable seizure outcome. Presence of VHFO may suggest a localized epileptogenic zone and predict favorable postoperative seizure outcome. Ictal VHFOs were 300

detected in a highly localized area (1 to 4 electrodes) and were generated just above the hyperintense regions on MRI. On the contrary, ictal HFOs sometimes occurred in regions beyond the epileptogenic zone to be resected for seizure relief. In all 7 patients with ictal VHFO, areas showing VHFO were resected completely. Ictal VHFO may be a more specific marker than HFO and SOZ for identification of the epileptogenic zone. The VHFOgenerating area may represent the “core” of the epileptogenic zone, although the boundary of the epileptogenic zone cannot be determined. If VHFOs are detected by a 10-kHz sampling recording, the area showing VHFO should be included in the resection. Detection of ictal VHFO may provide not only localizing information, but also prediction of seizure prognosis postsurgery. Limited resection of the VHFO-generating area may be sufficient for seizure relief. However, further studies are necessary to validate this hypothesis. Complete resection of areas showing HFO is not always necessary for seizure relief. HFO may not be so useful for determining the epileptogenic zone. HFOs may be a sensitive marker for epileptogenicity, but they are not specific for the epileptogenic zone. HFOs have been detected in nonepileptogenic functional cortices.7 Volume 78, No. 2

Usui et al: Significance of VHFO in epilepsy

SOZ was not identified in 3 of 6 patients with unfavorable outcome. In these patients, extent of surgical resection was determined by considering all other information, including the location of structural lesions and other noninvasive data. Poor identification of SOZ itself may be related to poor definition of the epileptogenic zone. Complete resection of the SOZ is not necessarily essential to achieve favorable outcome. On the other hand, 1 patient had unfavorable outcome even after complete resection of the SOZ. As already stated by other researchers, SOZ is not always a good marker of epileptogenic zone for resection.2,3 In 3 patients with incomplete MRI lesion resection, seizure outcome was unfavorable. The reason of incomplete resection in these patients was that the lesion was located in the close proximity to the eloquent cortex. It is well known that complete lesion resection is the most important factor predicting favorable seizure outcome. In all 7 patients with VHFO, histopathology revealed FCD type II. The electrodes recording VHFO were located just above the localized hyperintense regions on MRI. Abnormal synaptic connectivity and changes in neurotransmitter systems are believed to cause the high epileptogenicity of FCD.8–10 and may be mechanisms that lead to the generation of VHFO. Continuous epileptiform discharges are well known to be an EEG characteristic of FCD,11 although not specific. VHFO may also be an EEG characteristic of FCD type II and may directly reflect the mechanism of epileptogenesis in FCD. Surgical outcome of FCD type II has been reported to be favorable,12 and complete lesion resection is most important for seizure control. Presence of VHFO may suggest the presence of localized type II FCD. However, whether VHFOs are a specific EEG finding for FCD is not clear, given that there were only 3 patients with pathologies other than FCD in this study. Further studies including more patients with pathologies other than FCD are necessary. Regarding the mechanisms of VHFO generation, we speculate that low-amplitude VHFOs recorded by macroelectrodes may represent summated activities of multiple subgroups of neurons with various nonsynchronous firing rates and phases, given that individual neuronal firing rates do not reach the VHFO range.4 Waveforms of VHFO are not typical sinusoidal waves. Their interpeak intervals are not constant, and their amplitudes are variable even within 1 particular cluster. This may suggest that rather than synchronization oscillation, VHFOs represent overlapping neuronal activities arising from many nonsynchronized neuronal groups. However, the precise mechanism of VHFO generation has not yet been elucidated. Further studies using simulAugust 2015

taneous recording of EEG and single-unit recordings are necessary to clarify the mechanism of VHFO generation. Limitations This study included only 13 patients with intractable neocortical epilepsy. Further studies with large number of patients are necessary. Only 16 channels were monitored simultaneously owing to the limitation of the EEG machine. Although we selected 16 channels after completing and analyzing the conventional EEG recordings at a sampling rate of 200Hz, selection bias is unavoidable. Improvement of amplifiers that allow recording of a larger number of EEG channels is necessary. In this study, only ictal data were reviewed. We did not review interictal data, given that visual inspection is very timeconsuming and labor intensive. Automatic detection and spectral analysis of VHFO should be developed. Only 2 cases with no MRI abnormality were included in this study. Postoperative seizure outcome in patients with MRI-invisible lesions remains unfavorable. The usefulness of VHFO in MRI-invisible cases should be explored further. In conclusion, ictal VHFO may be a specific marker of the epileptogenic zone. The presence of VHFO correlates with favorable postsurgical seizure outcome. If VHFOs are detected, complete resection of the VHFO-generating area is mandatory for postoperative seizure relief.

Acknowledgment This study is supported by the Japan Epilepsy Research Foundation (N.U.).

Authorship N.U.: conception of the study, study design, carrying out the study and collecting data, analyzing the data, and writing the manuscript; K.T.: study design, collecting data, analyzing the data, and critically reviewing and editing the manuscript; K.B.: critically reviewing and editing the manuscript; K.M.: analyzing the pathological data, critically reviewing and editing the manuscript; K.U.: critically reviewing and editing the manuscript; T.T.: critically reviewing and editing the manuscript; T.M.: critically reviewing and editing the manuscript; Y.I.: critically reviewing and editing the manuscript.

References 1.

Ochi A, Otsubo H, Donner EJ, et al. Dynamic changes of ictal high-frequency oscillations in neocortical epilepsy: using multiple band frequency analysis. Epilepsia 2007;48:286–296.

301

ANNALS

of Neurology

2.

Jacobs J, Zijlmans M, Zelmann R, et al. High-frequency electroencephalographic oscillations correlate with outcome of epilepsy surgery. Ann Neurol 2010;67:209–220.

3.

Akiyama T, McCoy B, Go CY, et al. Focal resection of fast ripples on extraoperative intracranial EEG improves seizure outcome in pediatric epilepsy. Epilepsia 2011;52:1802–1811.

4.

Usui N, Terada K, Baba K, et al. Very high frequency oscillations (over 1000 Hz) in human epilepsy. Clin Neurophysiol 2010;121: 1825–1831.

5.

Jacobs J, Levan P, Chatillon CE, et al. High frequency oscillations in intracranial EEGs mark epileptogenicity rather than lesion type. Brain 2009;132:1022–1037.

6.

Otsubo H, Ochi A, Imai K, et al. High-frequency oscillations of ictal muscle activity and epileptogenic discharges on intracranial EEG in a temporal lobe epilepsy patient. Clin Neurophysiol 2008; 119:862–868.

7.

302

Nagasawa T, Juh asz C, Rothermel R, et al. Spontaneous and visually driven high-frequency oscillations in the occipital cortex: Intra-

cranial recording in epileptic patients. Hum Brain Mapp 2012;33: 569–583. 8.

Ferrer I, Pineda M, Tallada M, et al. Abnormal local-circuit neurons in epilepsia partialis continua associated with focal cortical dysplasia. Acta Neuropathol 1992;83:647–652.

9.

Mattia D, Olivier A, Avoli M. Seizure-like discharges recorded in human dysplastic neocortex maintained in vitro. Neurology 1995; 45:1391–1395.

10.

Spreafico R, Battaglia G, Arcelli P, et al. Cortical dysplasia: an immunocytochemical study of three patients. Neurology 1998;50: 27–36.

11.

Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37:476–487.

12.

Urbach H, Scheffler B, Heinrichsmeier T, et al. Focal cortical dysplasia of Taylor’s balloon cell type: a clinicopathological entity with characteristic neuroimaging and histopathological features, and favorable postsurgical outcome. Epilepsia 2002;43:33–40.

Volume 78, No. 2

Significance of Very-High-Frequency Oscillations (Over 1,000Hz) in Epilepsy.

We previously reported ictal very-high-frequency oscillations (VHFO) of 1,000 to 2,500Hz recorded by subdural macroelectrodes using a 10-kHz sampling ...
169KB Sizes 2 Downloads 10 Views