SEARCH

SEARCH BY CITATION

Keywords:

  • Functional near-infrared spectroscopy combined with electroencephalography;
  • Seizures;
  • Hemodynamic response;
  • Frontal lobe epilepsy

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

Purpose:  To investigate spatial and metabolic changes associated with frontal lobe seizures.

Methods:  Functional near-infrared spectroscopy combined with electroencephalography (EEG-fNIRS) recordings of patients with confirmed nonlesional refractory frontal lobe epilepsy (FLE).

Key Findings:  Eighteen seizures from nine patients (seven male, mean age 27 years, range 13–46 years) with drug-refractory FLE were captured during EEG-fNIRS recordings. All seizures were coupled with significant hemodynamic variations that were greater with electroclinical than with electrical seizures. fNIRS helped in the identification of seizures in three patients with more subtle ictal EEG abnormalities. Hemodynamic changes consisted of local increases in oxygenated (HbO) and total hemoglobin (HbT) but heterogeneous deoxygenated hemoglobin (HbR) behavior. Furthermore, rapid hemodynamic alterations were observed in the homologous contralateral region, even in the absence of obvious propagated epileptic activity. The extent of HbO activation adequately lateralized the epileptogenic side in the majority of patients.

Significance:  EEG-fNIRS reveals complex spatial and metabolic changes during focal frontal lobe seizures. Further characterization of these changes could improve seizure detection, localization, and understanding of the impact of focal seizures.

Functional near-infrared spectroscopy (fNIRS) is an emerging neuroimaging technology for continuous, non-invasive monitoring of oxygenated (HbO), deoxygenated (HbR) and total hemoglobin (HbT = HbO + HbR). Assuming constant hematocrit, changes in HbT can serve as indicators of cerebral blood volume (CBV) variations (Lloyd-Fox et al., 2010). According to the implementation technique described below, photons from two wavelengths in the near-infrared range (one more sensitive to HbO, the other to HbR) shone by optic fibers penetrate the skull and superficial cortex, are absorbed (mainly by HbO and HbR) and scattered or reflected back to the surface where they can be measured by photodetectors. Changes in HbO or HbR concentrations in the superficial cortex will cause alterations in reflected light intensity (Jöbsis, 1977). When fNIRS is combined with continuous electroencephalography (EEG), the hemodynamic changes occurring before, during, and after seizures can be investigated (Villringer et al., 1994; Steinhoff et al., 1996; Irani et al., 2007).

Combined EEG-fNIRS recently disclosed that temporal lobe seizures were associated with significant local and remote hemodynamic variations, which outlasted seizure duration (Nguyen et al., 2011). Complex temporal lobe seizures were coupled with an initial increase in HbT and HbO as well as a decrease in HbR, congruent with a compensatory increment of local CBV to augment oxygen supply to the epileptic area. This initial phase was, however, consistently followed by an increase in HbR, indicating that ictal metabolic demands are then unmet during seizure progression.

Frontal lobe epilepsy (FLE) is distinct from temporal lobe epilepsy in many ways. Clinically, the seizures are frequent but short, have a nocturnal tendency, include frequent motor signs, and are followed by little or no postictal confusion (Fogarasi et al., 2001). The extensive network of subcortical connections in the frontal lobes permits rapid and distant spread of seizure activity. Here, we report the hemodynamic observations made during continuous EEG-fNIRS recording of frontal lobe seizures.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

Patients

Patients with nonlesional refractory FLE, candidates for potential epilepsy surgery, were recruited for a session of simultaneous EEG-fNIRS recording at the Optical Imaging Laboratory of Hôpital Sainte-Justine. The study was approved by the Université de Montréal Hôpital Sainte-Justine and Hôpital Notre-Dame Ethics Committees, and informed consent was obtained from all subjects. To increase the likelihood of recording seizures, EEG-fNIRS studies were undertaken when patients were admitted for video-EEG monitoring as part of their presurgical evaluation, since anticonvulsants were then frequently tapered off. Scalp electrodes for clinical video-EEG monitoring were removed to install an EEG-fNIRS helmet and reinstalled after the recording session to resume video-EEG monitoring. An epileptologist was available at all times to ensure patient safety. After comprehensive, noninvasive, presurgical investigation, an intracranial study was performed to confirm epileptogenic zone localization.

Combined EEG-fNIRS recording

Methodologies for simultaneous EEG-fNIRS recordings and data analysis have been detailed previously in Nguyen et al. (2011). Briefly, EEG was recorded with a Neuroscan Synamps 2TM system (Compumedics, Charlotte, NC, U.S.A.) via 19 homemade carbon fiber electrodes inserted through a perforated helmet and fixed on the scalp according to the 10–20 system (500-Hz sampling rate; 0.1–100-Hz band-pass filter; 60-Hz notch filter). fNIRS measurements were acquired with a multichannel Imagent Tissue Oxymeter (ISS Inc., Champaign, IL, U.S.A.), using up to 64 fiber sources (operating at 690 nm and 830 nm, and modulated at a frequency of 110 MHz) and up to 16 photodetectors mounted on the helmet for optimal coverage of bilateral frontal regions as well as temporal and parietal regions. Later recordings had more channels than earlier ones as our technique improved over the years. The distance between the sources and detectors was kept between 3 and 5 cm. Except for Patient 1, there were no overlapping channels. Optical intensity (DC), modulation amplitude and phase changes in collected light data were sampled at 19.5 Hz. Only DC was used for hemodynamic analysis. Continuous, simultaneous EEG-fNIRS monitoring was carried out for 1–2 h. A camera filmed the whole session for offline review of ictal manifestations or artifact-generating movements. A pulse oximeter was attached to a fingertip for visual monitoring of arterial blood oxygen saturation. After the recording session, the location of every source, detector, and fiducial point was digitized and traced (Brainsight Frameless 39; Rogue Research, Montreal, QC, Canada) for coregistration (MATLAB, Mathworks, Natick, MA, U.S.A.) with the cortical surface (IMAGIC, Neuronic, Havana, Cuba).

Data analysis

After acquisition, the EEG data were reviewed offline by an epileptologist (DKN). Seizure-onset and seizure end-times were, respectively, defined as the earliest and latest clinical or electrographic evidence of seizure activity. Raw light intensity was visually inspected to detect periods where movement artifacts caused discontinuity of the signal so they could be rejected. In addition, channels were completely rejected when the distance between them was lower than 2.5 cm or higher than 5 cm, or if the standard deviation during baseline was higher than 20%. Changes in HbO and HbR were calculated from DC variations according to the modified Beer-Lambert law. The values reported do not presume to indicate exact concentrations in the brain, but rather reflect representative relative changes in the brain, as we used a uniform partial volume factor of 1. Concentration curves are presented as percentages of variation compared to typical steady-state values of HbT 100 μm, HbO 75 μm, and HbR 25 μm (Boas et al., 2003). Low-pass frequency, zero-phase digital filtering of the data was performed with a cutoff frequency of 0.5 Hz, reducing the fast cardiac signal component. Channels with raw DC at the level of equipment noise or with a standard deviation higher than 20% were considered as artifactual and excluded from the analysis. Optical seizure onset time was marked when the first obvious hemodynamic changes occurred and was compared with EEG seizure onset identified by the epileptologist. To track the average temporal evolution of hemoglobin variations on both sides of the epileptic focus (ipsilaterally and contralaterally) in the figures below, channels were selected and averaged, based on the epileptogenic zone as determined by intracranial EEG.

Topographic analysis

A topographic view was created to display the temporal and spatial evolution of HbO and HbR activations. A Student’s t-test was performed to ensure that the amplitude of the activation exceeded baseline noise using the following formula: t = X/SEM, where t is the uncorrected t-value, X is the amplitude of the concentration during the seizure and SEM is the standard error for 5 s of baseline. We mapped the t-value to a topographic view onto the skin segmentation using the patient’s magnetic resonance imaging (MRI). The activation was projected directly over the path between a source and a detector without interpolation using the optode coordinates acquired on the skin. When two channels passed through the same position, the maximal value was mapped.

Laterality indices

To evaluate if fNIRS could adequately lateralize the epileptic focus, two laterality indices were calculated. The extent laterality index (ELI), as a function of time, compares the extent of HbO activation between both hemispheres and is calculated as follows: ELI = (NL − NR)/(NL + NR), with NL = number of left channels and NR = number of right channels, keeping only channels exceeding 50% of the maximal t-value The peak laterality index (PLI) compares the peak t-value at every moment between both hemispheres and is calculated as follows: PLI (t) = [(max peak left (t) − max peak right (t)]/[(max peak left (t) + max peak right (t)].

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

Between June 2007 and July 2010, 20 patients with nonlesional, suspected refractory FLE underwent a continuous EEG-fNIRS study. Nine of these patients were excluded because they did not subsequently undergo an intracranial EEG study to confirm the location of the epileptogenic zone for the following reasons: pregnancy (1), evidence of multifocal seizures (1), drug abuse (1), psychosis (1), vagus nerve stimulation preferred to surgery (2), significant reduction in seizure frequency with additional drug changes (2), and excessive bleeding upon skull opening due to a coagulation disorder (1). Of the remaining 11 patients who subsequently underwent an intracranial study, all were found to have a frontal epileptogenic focus as initially suspected. Of these 11 subjects, at least one clear electrical or electroclinical seizure was identified during presurgical EEG-fNIRS recordings in 6 (Table 1, Patients 1, 3, 5–8). In 3 others (Table 1, Patients 2, 4, and 9), “probable” seizures were identified after careful review of EEG segments, guided by evidence of optical signal deflectance in fNIRS data. Patient 2 reported a brief seizure during a particular time-frame that coincided with a deflectance in the optical signal and associated with diffuse EEG de-synchronization; Patients 4 and 9 had no ictal manifestations, but the deflectance in the optical signal coincided with diffuse EEG de-synchronization with overlying low-voltage fast-activity followed by semirhythmic low-voltage slowing. Although less straightforward than for the first six patients, these ictal scalp EEG patterns were, however, congruent with ictal patterns observed during intracranial EEG recordings. In the two other implanted patients, no seizures occurred during EEG-fNIRS. Based on intracranial recordings, the nine identified patients (seven male, mean age 27 years, range 13–46 years) who presented seizures on EEG-fNIRS had an epileptogenic zone localized to the frontopolar cortex in 1, the mesial frontal region in 1, and the lateral frontal neocortex in the remainder (motor cortex in 1, dorsolateral premotor cortex in 1, and intermediate dorsolateral frontal region in 5).

Table 1.   Demographic data, site of epileptogenic zone based on intracerebral recordings, and date of EEG-fNIRS recordings with respective number of channels
PatientAge (years)SexEZDate of testNumber of fNIRS channels
110MR FP3/25/200744
213FR MC7/10/200761
345MR ILC (IFG)-Ins10/16/200896
439FL ILC (IFG)-Ins-STG11/5/200868
521MR ILC (IFG, MFG)2/12/2009140
621ML IMC (Medial FG, SFG), CC9/10/2009100
746ML DLPMC, ILC (SFG, MFG)11/13/2009203
813MR ILC (IFG)1/19/2010140
935MR ILC (IFG)4/16/2010142
Mean27   110

Local hemodynamic signal changes

A total of 18 seizures were identified: five complex partial seizures, five simple partial seizures, and eight electrical seizures (Table 2). Pulse oximetry monitoring indicated no significant peripheral desaturation during any seizure. Each ictal event was accompanied by changes in the NIRS optical signal over the epileptogenic zone. Ictal hemodynamic variations for each seizure are detailed in Figs 1–4 and S1–S9. On average, these changes occurred <2 s after the onset of EEG discharge (mean 1.39 ± 2.30 s). The optical signal alterations persisted after the end of EEG discharge in Patients 1, 2, 7, and 8. In the remainder, the duration of optical variations was relatively close to the duration of ictal EEG changes.

Table 2.   Details of hemodynamic changes in all recorded seizures: onset of fNIRS variations compared to first scalp EEG evidence of ictal activity, EEG evidence of seizure activity duration, duration of hemodynamic changes in individual seizures, peak concentration alterations recorded over the epileptogenic zone
PatientSzSymptomsDelay to fNIRS changes (s)Sz duration EEG (s)Sz duration fNIRS (s)PeakFigures
HbTTimeHbOTimeHbRTime
  1. HbT, total hemoglobin; HbO, oxyhemoglobin; HbR, deoxyhemoglobin; SD, standard deviation.

  2. *estimated (limited by artefact).

11SPS: sudden fear335506.83299.3627−3.0124See Gallagher et al. (2008)
2SPS: sudden fear836508.012010.08201.6720
3ES516201.32111.8211−0.1811
4ES36100.8051.0650.015
21SPS: shock in L arm0371.473.22.083−0.473Figure S1
31ES013161.14111.4211−0.288 Figure 1
2ES012130.9881.348−0.4413 Figure 1
41ES0661.083.021.473−1.508 Figure 2
2ES112.26150.516.830.947−0.817Figure S2
51CPS: gelastic sz062  62*3.73116.4112*−6.09* 13*Figure S3
61CPS: arousal016 16*11.031613.2164.516Figure S4
2CPS: arousal414 15*7.68148.90144.0214Figure S5
71SPS: jaw + arm jerks020305.12154.69157.0215 Figure 3
2SPS: subtle hand jerks14101.3141.1841.724Figure S6
81CPS: fixed gaze024357.361410.815−11.6529 Figure 4
2CPS: fixed gaze038   38*4.32216.8721−10.39*  26*Figure S7
91ES04.8150.652.940.7530.433Figure S8
2ES04.8250.603.440.7530.195Figure S9
Mean total  1.3918.2319.503.35 4.34 −0.98  
SD total  2.3015.6416.303.33 4.21 5.26  
Electrical seizures versus clinical seizures
Mean ES   9.4211.250.88 1.19 −0.32  
SD total   4.465.650.27 0.38 0.61  
Mean CS   25.2831.305.68 7.36 −1.36  
Total SD   17.9618.993.07 3.86 6.37  
image

Figure 1.   (A) Channel configuration. (B) Ictal hemodynamic variations (HbT, black; HbO, red; HbR, blue) over the epileptogenic zone (closed circles) and the contralateral homologous region (open circles). The black line indicates time of EEG evidence of seizure activity. (C) Topographic uncorrected t-statistic viewed at different time points (black triangles) during the course of the seizure. The epileptogenic zone is indicated by a circle. This 45-year-old man (Patient 3) with right frontoinsular epilepsy had two electrical seizures back to back during EEG-fNIRS recording associated each time with an increase in HbT and HbO over the epileptogenic zone and in the contralateral homologous region.

Download figure to PowerPoint

image

Figure 2.   Same legend as in Fig. 1. This 39-year-old woman (Patient 4) with left perisylvian epilepsy did not experience any clinical seizures during EEG-fNIRS. A first look at the EEG failed to identify obvious electrical seizures. However, review of the fNIRS signal revealed HbO and HbT increases (and HbR decrease) at a certain time period. Closer attention to that particular EEG time frame disclosed a previously missed electrical seizure characterized by EEG de-synchronization with overlying, low-voltage fast-activity followed by semirhythmic low-voltage slowing.

Download figure to PowerPoint

image

Figure 3.   Same legend as in Fig. 1. This 46-year-old man (Patient 7), with a left premotor cortex focus, experienced clonic jerks of the right jaw and arm. Although each individual jerk generated a small artifact (arrow), the fNIRS signal in between jerks was of good quality. Note how this electroclinical seizure led to higher HbT, HbO, and HbR variations compared to electrical seizures shown in Figs. 1 and 2. Also note the HbR increase despite the HbT and HbO increment during the ictus.

Download figure to PowerPoint

image

Figure 4.   Same legend as in Fig. 1. This 13-year-old man (Patient 8) had seizures originating from the right inferior frontal gyrus. Contrary to the seizure shown in Fig. 3 (Patient 7), this seizure was associated with a significant HbR decrease during ictus.

Download figure to PowerPoint

Over the epileptogenic zone, HbO and HbT followed a bell-shaped curve with an initial rise from baseline to peak, followed by a gradual decline to a plateau or to initial baseline (see part B of Figs. 1–4 and S1–S9). The mean maximal amplitude of HbT and HbO increase was 3.35 ± 3.33% and 4.34 ± 4.21% from baseline, respectively (Table 2). Electroclinical seizures led to higher increments of HbT and HbO (5.68 ± 3.07% and 7.36 ± 3.86%, respectively) compared to electrical seizures (0.88 ± 0.27 and 1.19 ± 0.38%, respectively). The peak amplitude of HbO increase was strongly correlated with seizure duration (r = 0.56). The behavior of HbR changes was more heterogeneous (see part B of Figs. 1–4 and S1–S9). Although an inverted bell-curve (initial decrease to a nadir, followed by a progressive return to baseline or plateau) was noted in four (Patients 1, 4, 5, and 8), an opposite response was observed in Patients 6 and 7, with very little change in the remaining subjects. As with HbO, the nadir of HbR decrease correlated with seizure duration (r = −0.46).

Remote hemodynamic changes

In addition to the local hemodynamic alterations described above, significant changes were also noted in the homologous contralateral region (see part C of Figs. 1–4 and S1–S9) where HbT, HbO and HbR behavior closely mirrored that over the epileptogenic zone (see Figs. 1–4 and S1–S9). No clear delay between the timing of ipsilateral versus contralateral changes was noted. The amplitude of HbO increase during the first seconds after seizure onset was slightly higher ipsilaterally in 8 (67%) of the 12 seizures and peaked slightly higher ipsilaterally in half of them. Of the five seizures (from Patients 3, 4, and 6) associated with a decrease in HbR, the highest nadir was noted ipsilaterally in 2 (40%), contralaterally in 1 (20%), and bilaterally in the remainder.

Lateralization

To evaluate if fNIRS could adequately lateralize the epileptic focus, two laterality indices were calculated for eight of the nine patients (patient 1 did not have contralateral coverage) (Fig. 5). At every time point, the ELI compares the extent of HbO activation between both hemispheres, whereas the PLI compares the peak t-value at every moment between both hemispheres. In the early phase of seizures, the ELI adequately identified the epileptogenic hemisphere in 7 (88%) of 8 patients. The PLI adequately identified the epileptogenic hemisphere in only 50% of patients (Table 3).

image

Figure 5.   The laterality index of HbO variation in function of time during seizure(s) in each patient (except Patient 1, who did not have bilateral coverage). The data were averaged when there was more than one seizure per patient. Blue: Extent laterality index (ELI) with a half maximum t-value threshold. Red: Maximal peak laterality index (PLI). The green line indicates the correct side of the epileptogenic zone for each patient.

Download figure to PowerPoint

Table 3.   Lateralization indexes: Time LI = time when index of laterality was highest
PatientSzTime LIELIConcordant with side of focusPLIConcordant with side of focus
  1. ELI, extent laterality index; PLI, peak laterality index.

  2. A value between −1 and −0.1 indicates lateralization to the R hemisphere, whereas a value between 0.1 and 1 indicates lateralization to the L hemisphere.

213−0.34Yes−0.30Yes
316−0.35Yes−0.12No
226−0.53 −0.039 
413.50.20Yes−0.07No
241 0.38 
513−1Yes−0.33Yes
6171Yes0.41Yes
270.27 0.18 
7130.67Yes0No
250.55 0.12 
815−1No−0.46No
250.37 0.28 
913−0.76Yes−0.16Yes
23−0.22 −0.10 
Percentage of patient adequately lateralized  88% 50%

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

In this study, we employed a noninvasive method to measure cortical hemodynamic and metabolic changes occurring throughout seizures, with high temporal resolution. We showed that focal frontal lobe seizures are associated with complex hemodynamic variations, that is, a local, expected increase in HbO and HbT but heterogeneous HbR behavior and, surprisingly, very early activation of the homologous region contralateral to the epileptogenic zone.

The observed ictal increments of HbT and HbO over the epileptogenic zone are congruent with the brain’s attempt to perfuse active, epileptic neurons with oxygenated hemoglobin by elevating local cerebral blood flow (CBF) and CBV (Suh et al., 2006). In our study, brief electrical seizures (inline image = 9.4 s) were associated with only small variations from baseline (inline image = 0.88% for HbT and 1.19% for HbO), but longer (inline image = 25.3 s) electroclinical seizures led to HbT and HbO increases as high as 11.0% and 13.2%, respectively (inline image = 5.68% for HbT and 7.36% for HbO). In a similar investigation of temporal lobe seizures (Nguyen et al., 2011), our group recently reported even higher changes from baseline (inline image = 11% for HbT and HbR), most likely due to their lengthier duration (inline image = 77.3 s). The longest temporal lobe complex partial seizure recorded (106 s) resulted in the highest HbT (17%) and HbO (18%) variations from baseline. These variations were three to eight times larger than those observed during normal cognitive processing (Gallagher et al., 2007), indicating that seizures are an abnormal physiologic state placing supranormal demands on the brain’s autoregulatory mechanisms.

Some controversy remains in the literature as to whether local CBF and CBV increases are adequate to meet these supranormal demands throughout the ictus (Suh et al., 2006). Although the inadequacy of CBF in addressing metabolic demands has been demonstrated in animal models of status epilepticus (Meldrum & Brierly, 1973; Meldrum & Horton, 1973; Kreisman et al., 1984), there is growing evidence of inadequate oxygenation at the onset or throughout shorter-duration ictal events as well (Bahar et al., 2006; Zhao et al., 2007, 2009). In our study of complex partial temporal lobe seizures (Nguyen et al., 2011), HbR (after an initial decrease) rose consistently as the seizure progressed, indicating that metabolic demands of the seizing tissue were not sufficiently compensated by the reactive regional increase in cerebral blood supply to wash out HbR. In this series of frontal lobe seizures, the ictal increments in CBF and CBV seemed sufficient for the majority of seizures, since HbR showed little change (suggesting adequate HbO supply) or decreased (indicating HbO oversupply) for most of them. Although this could be ascribed to longer seizure duration in our temporal lobe epilepsy series compared to the present FLE series, other factors are probably in play, as exemplified by Patient 6, who presented an increase in HbR (indicating an insufficient neurovascular response) for a “mere” 20-s seizure. In rare EEG-fMRI analyses of fortuitous motionless seizures, significant generally positive blood oxygen level dependent (BOLD) changes (congruent with a decrease in HbR) have been reported in the seizure-onset zone along with other remote smaller clusters of BOLD signal change (Kobayashi et al., 2006a,b; Donaire et al., 2009; Tyvaert et al., 2009; LeVan et al., 2010; Thornton et al., 2010; Chaudhary et al., 2011). However, additional areas of BOLD signal decrease during seizures have been observed within the presumed seizure onset zone, in surrounding areas of BOLD signal increase, remotely from the seizure onset zone including on the opposite hemisphere. Although some have been ascribed to the default mode network, the significance of these BOLD signal decreases remains largely unknown. Further work is necessary to better ascertain the different potential factors underlying heterogeneity in the HbR response: seizure duration, location of seizure onset, volume of epileptogenic tissue involved during ictus, seizure frequency, underlying substrate of epileptogenicity, age, comorbid medical conditions, and so on (Sokol et al., 2000; Haginoya et al., 2002; Shuhaiber et al., 2004).

In addition to hemodynamic changes over the epileptogenic zone described earlier, our study reveals significant modulations in contralateral homologous regions, which almost mirrored those seen ipsilaterally. This is in line with some prior observations both on interictal spikes and focal seizures. Using intrinsic optical imaging, Schwartz and Bonhoeffer (2001) showed a clear increase in the optical signal in the contralateral hemisphere, homotopic to the acute bicuculline spiking foci in the ferret somatosensory cortex. The authors ascribed these contralateral changes to orthodromic effects of transcallosal projected neuronal activity (Schwartzkroin et al., 1975). In animal and human EEG-fMRI studies, widespread and distant (both positive and negative) BOLD signal changes are seen with focal interictal epileptic spikes or brief seizures (Kobayashi et al., 2006a; Englot et al., 2008; Truccolo et al., 2011). Using EEG spectral analysis, Yu et al. (2009) revealed significant contralateral EEG changes (predominating in lower frequencies) in 90% of contralateral BOLD activations triggered by spikes. These authors suggested that these spectral changes in areas corresponding to contralateral activations possibly reflected poorly synchronized but intense neuronal activity. It has been known for some time that frontal lobe seizure activity can spread rapidly through various pathways leading to widespread regional or multilobar interictal and ictal signal abnormalities both on scalp and intracranial EEG (Quesney et al., 1992; Salanova et al., 1993). More recently, however, it has been argued that rapid discharges at seizure onset observed over different brain sites were not necessarily due to propagation but instead to quasi-synchronous involvement of a network of neuronal populations distributed in distinct and distant brain structures (Bartolomei et al., 2008; Wendling et al., 2010). In our study, the limitations of scalp EEG (low signal to noise ratio, inability to detect early ictal rhythms emanating from deeper structures, attenuation or cancellation of signal by soft tissues and bone, signal artifacts, and so on) and the limited number of seizures per patient prevent us from any clear conclusions on the pathophysiology of these contralateral and remote hemodynamic changes.

Prior studies have suggested that EEG-fNIRS could be used to localize the epileptogenic zone due to its high temporal resolution (Adelson et al., 1999; Watanabe et al., 2000; Schichiri et al., 2001; Watanabe et al., 2002; Gallagher et al., 2008), in contrast to ictal single photon computed tomography, for example, which can only provide a single snapshot of CBF several seconds after seizure onset when epileptic activity has already propagated to other regions (Goffin et al., 2008). These studies were, however, limited by the small number of optodes used. Benefiting from much larger spatial sampling, our demonstration of rapid and significant contralateral hemodynamic changes, even before evidence of seizure propagation, complicates the issue of lateralization/localization. In this preliminary work, the extent of HbO activation during the early phase of seizures was more indicative of the epileptogenic hemisphere than peak HbO activation. Additional work is required to establish the localization value of fNIRS. Can mirror activations be exploited to confirm the location of the focus (Huberfeld et al., 2006)? Can the presence of a mirror image facilitate the identification of the true epileptic focus in the presence of several areas of activation? Can the analysis of the preictal period be used for seizure localization (Zhao et al., 2007; Tyvaert et al., 2009)? Putting aside its localization potential, fNIRS was found to be a robust tool for seizure detection in our study, as analyzing the optical signal was able to guide a more careful review of a specific time-frame of the EEG signal to detect seizures with subtle electrographic changes.

Limitations

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

One disadvantage of EEG-fNIRS is that only superficial neocortical hemodynamic changes can be assessed. Fortunately, the majority of our patients in this series had a neocortical frontal lobe focus. Because frontal lobe seizures are frequently associated with motor symptoms, fNIRS data acquisition could have been disturbed by movement during seizures. Fortunately, most recorded seizures generated no movements (electrical seizures) or only subtle or small movements (during electroclinical seizures). Furthermore, we were careful to exclude channels exhibiting aberrant signals. It should also be remembered that larger activation areas will also induce larger fNIRS responses due to partial volume effects. Untangling spatial effects from increased local demands remains difficult and will require tomographic approaches. Finally, the influence of extracerebral tissue (skin and bone hemoglobin) on cerebral fNIRS signals in our study is unknown (Okada & Delpy, 2003). Our setting, which favored large spatial sampling, did not include short-source detector separation channels to allow subtraction of contamination from superficial layers.

Conclusion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

Noninvasive, continuous EEG-fNIRS recording of frontal lobe seizures with large spatial sampling reveals HbT and HbO increases but heterogeneous HbR responses, not only over the epileptogenic zone but also in the contralateral homologous region. Further work is necessary to elucidate the pathophysiology underlying these observations.

Acknowledgments

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

This work was supported by Fonds de la recherche en santé du Québec (FRSQ) Grant 14385, the Canadian Institutes of Health Research (CIHR) Institute of Circulatory and Respiratory Health (ICRH) and the Heart and Stroke Foundation of Canada (HSFC) Grant 62573, and the Savoy Foundation.

Disclosure

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

All authors report no disclosures. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information
  • Adelson PD, Nemoto E, Scheuer M, Painter M, Morgan J, Yonas H. (1999) Noninvasive continuous monitoring of cerebral oxygenation periictally using near-infrared spectroscopy: a preliminary report. Epilepsia40:14841489.
  • Bahar S, Suh M, Zhao M, Schwartz TH. (2006) Intrinsic optical signal imaging of neocortical seizures: the ‘epileptic dip’. Neuroreport17:499503.
  • Bartolomei F, Chauvel P, Wendling F. (2008) Epileptogenicity of brain structures in human temporal lobe epilepsy: a quantified study from intracerebral EEG. Brain131:18181830.
  • Boas DA, Strangman G, Culver JP, Hoge RD, Jasdzewski G, Poldrack RA, Rosen BR, Mandeville JB. (2003) Can the cerebral metabolic rate of oxygen be estimated with near-infrared spectroscopy?Phys Med Biol48:24052418.
  • Chaudhary UJ, Duncan JS, Lemieux L. (2011) Mapping hemodynamic correlates of seizures using fMRI: a review. Hum Brain Mapp. Doi: 10.1002/hbm.21448.
  • Donaire A, Bargallo N, Falcon C, Maestro I, Carreno M, Setoain J, Rumia J, Fernandez S, Pintor L, Boget T. (2009) Identifying the structures involved in seizure generation using sequential analysis of ictal-fMRI data. Neuroimage47:173183.
  • Englot DJ, Mishra AM, Mansuripur PK, Herman P, Hyder F, Blumenfeld H. (2008) Remote effects of focal hippocampal seizures on the rat neocortex. J Neurosci28:90669081.
  • Fogarasi A, Janszky J, Faveret E, Pieper T, Tuxhorn I. (2001) A detailed analysis of frontal lobe seizure semiology in children younger than 7 years. Epilepsia42:8085.
  • Gallagher A, Thériault M, Maclin E, Low K, Gratton G, Fabiani M, Gagnon L, Valois K, Rouleau I, Sauerwein HC, Carmant L, Nguyen DK, Lortie A, Lepore F, Béland R, Lassonde M. (2007) Near-infrared spectroscopy (NIRS) as an alternative to the Wada test for language mapping in children, adults and special populations. Epileptic Disord9:241255.
  • Gallagher A, Lassonde M, Bastien D, Vannasing P, Lesage F, Grova C, Bouthillier A, Carmant L, Lepore F, Béland R, Nguyen DK. (2008) A non-invasive pre-surgical investigation in a 10-year-old epileptic boy using simultaneous NIRS-EEG. Seizure17:576582.
  • Goffin K, Dedeurwaerdere S, Laere KV, Van Paescchen W. (2008) Neuronuclear assessment of patients with epilepsy. Semin Nucl Med38:227239.
  • Haginoya K, Munakata M, Kato R, Yokoyama H, Ishizuka M, Inuma K. (2002) Ictal cerebral haemodynamics of childhood epilepsy measured with near-infrared spectrophotometry. Brain125:19601971.
  • Huberfeld G, Habert MO, Clemenceau S, Maksud P, Baulac M, Adam C. (2006) Ictal brain hyperperfusion contralateral to seizure onset: the SPECT mirror image. Epilepsia47:123133.
  • Irani F, Platek SM, Bunce S, Ruocco AC, Chute D. (2007) Functional near infrared spectroscopy (fNIRS): an emerging neuroimaging technology with important applications for the study of brain disorders. Clin Neuropsychol21:937.
  • Jöbsis FF. (1977) Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science198:12641267.
  • Kobayashi E, Bagshaw AP, Benar CG, Aghakhani Y, Andermann F, Dubeau F, Gotman J. (2006a) Temporal and extratemporal BOLD responses to temporal lobe interictal spikes. Epilepsia47:343354.
  • Kobayashi E, Hawco CS, Grova C, Dubeau F, Gotman J. (2006b) Widespread and intense BOLD changes during brief focal electrographic seizures. Neurology66:10491055.
  • Kreisman NR, Sick TJ, Rosenthal M. (1984) Concepts of brain oxygen sufficiency during seizures. Adv Exp Med Biol180:381392.
  • LeVan P, Tyvaert L, Moeller F, Gotman J. (2010) Independent component analysis reveals dynamic ictal BOLD responses in EEG-fMRI data from focal epilepsy patients. Neuroimage49:366378.
  • Lloyd-Fox S, Blasi A, Elwell CE. (2010) Illuminating the developing brain: the past, present and future of functional near infrared spectroscopy (review). Neurosci Biobehav Rev34:269284.
  • Meldrum BS, Brierly JB. (1973) Prolonged epileptic seizures in primates: ischemic cell change and its relation to ictal physiological events. Arch Neurol28:1017.
  • Meldrum BS, Horton RW. (1973) Physiology of status epilepticus in primates. Arch Neurol28:19.
  • Nguyen DK, Tremblay J, Pouliot P, Vannasing P, Florea O, Carmant L, Lepore F, Sawan M, Lesage F, Lassonde M. (2011) Non-invasive continuous EEG-fNIRS recording of temporal lobe seizures. Epilepsy Res99:112126.
  • Okada E, Delpy DT. (2003) Near-infrared light propagation in an adult head model. II. Effect of superficial tissue thickness on the sensitivity of the near-infrared spectroscopy signal. Appl Opt42:29152922.
  • Quesney LF, Constain M, Rasmussen T, Olivier A, Palmini A. (1992) Presurgical EEG investigation in frontal lobe epilepsy. Epilepsy Res5:5569.
  • Salanova V, Morris HH III, Van Ness PC, Luders H, Dinner D, Wyllie E. (1993) Comparison of scalp electroencephalogram with subdural electrocorticogram recordings and functional mapping in frontal lobe epilepsy. Arch Neurol50:294299.
  • Schichiri M, Tanabe T, Hara K, Suzuki S, Wakamiya E, Tamai H. (2001) Usefulness of near-infrared spectroscopy for identification of epileptic foci in two localization-related epilepsy patients. No To Hattatsu33:475479.
  • Schwartz TH, Bonhoeffer T. (2001) In vivo optical mapping of epileptic foci and surround inhibition in ferret cerebral cortex. Nat Med7:10631067.
  • Schwartzkroin PA, Futamachi KJ, Noebels JL, Prince DA. (1975) Transcallosal effects of a cortical epileptiform focus. Brain Res99:5968.
  • Shuhaiber H, Bolton S, Alfonso I. (2004) Cerebral regional oxygen fluctuations and decline during clinically silent focal electroencephalographic seizures in a neonate. J Child Neurol19:539540.
  • Sokol DK, Markand ON, Daly EC, Luerssen TG, Malkoff MD. (2000) Near infrared spectroscopy (NIRS) distinguishes seizure types. Seizure9:323327.
  • Steinhoff BJ, Herrendorf G, Kruth C. (1996) Ictal near infrared spectroscopy in temporal lobe epilepsy: a pilot study. Seizure5:97101.
  • Suh M, Ma H, Zhao M, Sharif S, Schwartz TH. (2006) Neurovascular coupling and oximetry during epileptic events. Mol Neurobiol33:181197.
  • Thornton RC, Rodionov R, Laufs H, Vulliemoz S, Vaudano A, Carmichael D, Cannadathu S, Guye M, McEvoy A, Lhatoo S, Bartolomei F, Chauvel F, Chauvel P, Diehl B, De Martino F, Elwes RDC, Walker MC, Duncan JS, Lemieux L. (2010) Imaging haemodynamic changes related to seizures: comparison of EEG-based general linear model, independent component analysis of fMRI and intracranial EEG. Neuroimage53:196205.
  • Truccolo W, Donoghue JA, Hochberg LR, Eskandar EN, Madsen JR, Anderson WS, Brown EN, Halgren E, Cash SS. (2011) Single-neuron dynamics in human focal epilepsy. Nat Neurosci14:635641.
  • Tyvaert L, LeVan P, Dubeau F, Gotman J. (2009) Noninvasive dynamic imaging of seizures in epileptic patients. Hum Brain Mapp30:39934011.
  • Villringer A, Planck J, Stodieck S, Boetzel K, Schleinkofer L, Dirnagl U. (1994) Noninvasive assessment of cerebral haemodynamics and tissue oxygenation during activation of brain cell function in human adults using near infrared spectroscopy. Adv Exp Med Biol345:559565.
  • Watanabe E, Maki A, Kawaguchi F, Mayanagi Y. (2000) Noninvasive cerebral blood volume measurement during seizures using multichannel near infrared spectroscopic topography. J Biomed Opt5:287290.
  • Watanabe E, Nagahori Y, Mayanagi Y. (2002) Focus diagnosis of epilepsy using near-infrared spectroscopy. Epilepsia43(Suppl. 9):5055.
  • Wendling F, Chauvel P, Biraben A, Bartolomei F. (2010) From intracerebral EEG signals to brain connectivity: identification of epileptogenic networks in partial epilepsy. Front Syst Neurosci4:154.
  • Yu JM, Tyvaert L, Levan P, Zelmann R, Dubeau F, Gotman J, Kobayashi E. (2009) EEG spectral changes underlying BOLD responses contralateral to spikes in patients with focal epilepsy. Epilepsia50:18041809.
  • Zhao M, Suh M, Ma H, Perry C, Geneslaw A, Schwartz TH. (2007) Focal increases in perfusion and decreases in hemoglobin oxygenation precede seizure onset in spontaneous human epilepsy. Epilepsia48:20592067.
  • Zhao M, Ma H, Suh M, Schwartz TH. (2009) Spatiotemporal dynamics of perfusion and oximetry during ictal discharges in the rat neocortex. J Neurosci29:28142823.

Supporting Information

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusion
  8. Acknowledgments
  9. Disclosure
  10. References
  11. Supporting Information

Figure S1. (A) Channel configuration. (B) Ictalhaemodynamic variations (HbT –black, HbO –red, HbR–blue) over the epileptogenic zone (closed circles) and thecontralateral homologous region (open circles). The black lineindicates time of EEG evidence of seizure activity. (C) Topographicuncorrected T-stats viewed at different time points (blacktriangles) during the course of the seizure.

Figure S2. Second seizure identified during EEG-fNIRS recording of Patient 4.

Figure S3. This 21-year-old male (Patient 5), with a focus in the posterior portion of the right inferior/middle frontal gyri, experienced a gelastic seizure.

Figure S4. This 21-year-old male (Patient 6) had a left medial frontal focus extending to the anterior portion of the superior frontal gyrus.

Figure S5. Second seizure recorded for Patient 6.

Figure S6. Second seizure recorded for Patient 7.

Figure S7. Second seizure recorded for Patient 8.

Figure S8. As in Patient 4, this 35-year-old male (Patient 9) was found to have an electrical seizure upon closer inspection of a certain EEG timeframe guided by the occurrence of fNIRS signal deflectance.

Figure S9. Second seizure recorded for Patient 9.

Table S1. Results of other functional tests performed in the presurgical evaluation: single photon emission computed tomography (SPECT), positron emission tomography (PET) and neuropsychological evaluation.

FilenameFormatSizeDescription
epi12011_sm_FigS1.tif5929KSupporting info item
epi12011_sm_FigS2.tif4816KSupporting info item
epi12011_sm_FigS3.tif4721KSupporting info item
epi12011_sm_FigS4.tif3183KSupporting info item
epi12011_sm_FigS5.tif4151KSupporting info item
epi12011_sm_FigS6.tif5499KSupporting info item
epi12011_sm_FigS7.tif5195KSupporting info item
epi12011_sm_FigS8.tif5174KSupporting info item
epi12011_sm_FigS9.tif3970KSupporting info item
epi12011_sm_TableS1.xls14KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.