1. The document discusses whether ictal EEG source imaging is feasible and reliable for pre-surgical evaluation of epilepsy patients. Several studies have explored different strategies for performing ictal EEG source imaging, including imaging spikes during seizures, imaging specific frequency bands, and combining source imaging with functional connectivity analysis.
2. One prospective study found that ictal source imaging was possible in 31% of patients and provided key information for surgical decision making in 14% of patients. Sensitivity was 70% and specificity was 76%.
3. Additional studies have found concordance between ictal source imaging and invasive monitoring in some cases. Combining source imaging with functional connectivity analysis has also shown potential for localizing the seizure onset
1. Ictal EEG source imaging:
is it the future?
P. van Mierlo, W. Staljanssens, G. Strobbe, V. Keereman, G. Birot, A Meurs,
E. Carrette, D. Van Roost, K. Vonck, P. Boon, M. Seeck, S. Vulliémoz
3. Problem statement
EEG Source Imaging has been used primarily on interictal events
because
higher SNR
less contamination by artefacts
more likely to be captured in recordings of limited duration
QUESTION: is ictal EEG Source Imaging feasible and
reliable to be used in the pre-surgical evaluation?
4. Ictal EEG source imaging strategies
1. ESI of spikes during the seizure
2. ESI at a certian frequency band of interest
3. Combine ESI and functional connectivity to find the driver of the
seizure
5. • Prospective study: EEG recorded in 100 patients
• In 93 patients ictal activity recorded, in 31 ictal ESI possible
Ictal Source Localization in Presurgical Patients With Refractory Epilepsy
*Paul Boon, *Michel D’Havé, *§Bart Vanrumste, *§Gert Van Hoey, *Kristl Vonck,
*Phyllis Van Walleghem, †Jacques Caemaert, ‡Eric Achten, and *Jacques De Reuck
Journal of Clinical Neurophysiology 19(5):461–468, 2002
bilateral independent frontotemporal spikes;
four complex partial seizures bilateral recruitment.
Optimum MRI: left hippocampal atrophy .
FDG PET: widespread left hemisphere hypometabolism.
modified left-side temporal lobectomy and hippocampectomy without further
invasive recording. The patient has been seizure free for more than 1 year.
6. In 14% of patients, it proved to be a key element in the surgical
decision process
Ictal Source Localization in Presurgical Patients With Refractory Epilepsy
*Paul Boon, *Michel D’Havé, *§Bart Vanrumste, *§Gert Van Hoey, *Kristl Vonck,
*Phyllis Van Walleghem, †Jacques Caemaert, ‡Eric Achten, and *Jacques De Reuck
Journal of Clinical Neurophysiology 19(5):461–468, 2002
Video-EEG: bilateral frontotemporal spikes; three complex partial
seizures: right frontotemporal recruitment.
Optimum MRI: left hippocampal atrophy.
FDG-positron emission tomography: right temporal hypometabolism.
Surgical decision: not to proceed with invasive recording;
drop out of presurgical evaluation.
7.
8. The ictal source localization had a sensitivity of 70% and a specificity of 76%. The positive
predictive value (PPV) for seizure freedom was 92% and the negative predictive value
(NPV) was 43%.
9. High-density ESI localized ictal onsets to the same region as intracranial
monitoring in 8 of 10 cases
High-density ESI has the potential to assist in the noninvasive
localization of epileptic seizures and to guide the placement of invasive
electrodes for localizing seizure onset.
Limitation: template head models were used
10. Ictal EEG source imaging strategies
1. ESI of spikes during the seizure
2. ESI at frequency band of interest
3. Combine ESI and functional connectivity to find the driver of the
seizure
11. 14 focal epilepsy patients that underwent long-term high density EEG
monitoring at Geneva Hospital.
Ictal epoch of 2s was selected and a narrow band-pass filter around the
fundamental seizure frequency was applied.
ESI was done using Locally Spherical Model with Anatomical
Constraints and LORETA as inverse technique
Comparisson with resection and interictal ESI was done
12.
13. 9 of 14 patients, interictal and ictal
ESI solutions were concordant
In 4 patient more than 1 spike
cluster
In 1 patient discordant
14.
15. Periodic Waveform Analyse
Beamformer
Head model
Ictal Source Localization
EEG
• Find most dominant early rhythmic
pattern
• frequency dependent time window
which had to contain at least eight ictal
waves or discharges
3D activity in the brain
Gritsch G et al. Automatic detection of the seizure onset zone based on ictal EEG.ConfProc IEEE EngMedBiol Soc. 3901-4. (2011)
Johannes Koren, Gerhard Gritsch, Gregor Kasprian, Christoph Baumgartner
Karl Landsteiner Institut für klinische Epilepsieforschung – Rosenhügel, Wien, Austrian Institute
of Technology (AIT), Wien, Universitätsklinik für Neuroradiologie, AKH Wien
16. Johannes Koren, Gerhard Gritsch, Gregor Kasprian, Christoph Baumgartner
Karl Landsteiner Institut für klinische Epilepsieforschung – Rosenhügel, Wien, Austrian Institute
of Technology (AIT), Wien, Universitätsklinik für Neuroradiologie, AKH Wien
17. Sensitivity 92%
Specificity 60%
PPV 67%
NPV 90%
These performance measures
were statistical significant
(p=0.013)
Koren J et al. Automatic ictal source localization in presurgical epilepsy evaluation (2017) under review
Johannes Koren, Gerhard Gritsch, Gregor Kasprian, Christoph Baumgartner
Karl Landsteiner Institut für klinische Epilepsieforschung – Rosenhügel, Wien, Austrian Institute
of Technology (AIT), Wien, Universitätsklinik für Neuroradiologie, AKH Wien
18. Beyond the Double Banana: Improved Recognition of Temporal Lobe Seizures in Long-
Term EEG
Ivana Rosenzweig,*† András Fogarasi,‡ Birger Johnsen,§ Jørgen Alving,* Martin Ejler
Fabricius,k Michael Scherg,¶ Miri Y. Neufeld,# Ronit Pressler,** Troels W. Kjaer,†† Walter
van Emde Boas,‡‡ and Sándor Beniczky*§,
Journal of Clinical Neurophysiology Volume 31, Number 1, February 2014
Phase maps had the highest sensitivity (+20%) and identified ictal activity at earlier
time-point than visual inspection
19. Ictal EEG source imaging strategies
1. ESI of spikes during the seizure
2. ESI at a certian frequency band of interest
3. Combine ESI and functional connectivity to find the driver of the
seizure
22. Method: ESI power vs EEG source connectivity
Ictal epoch selection EEG source imaging Multiple active sources
L R P A
Source 1 Source 3 Source 5
Source 2 Source 4 Source 6
0 1 2 3 4 5 6
1
2
3
4
5
6
Time (s)
Source
Comparison with resectionSeizure onset zone localization
Power
Connectivity
LE connectivity
LE power
Resection
23. EEG source connectivity vs power: hd-EEG
EEG source connectity
Staljanssens et al., Brain Topography 2016
24. Patients
Mean age epilepsy onset: 19
Mean age epilepsy surgery: 34
Inclusion criteria:
• patients who underwent presurgical evaluation
• one-time resective epilepsy surgery
• structural MRI before and after resection
• Engel class 1 with minimum follow-up of 1 year post-operatively
Surgery type:
amygdalo-hippocampectomy: 10
partial lobectomy: 10
tailored lesionectomy: 5
topectomy: 1
Mean follow-up: 3.8 years
111 seizures in 27 patients:
No clear discharges
Right rhythmic
Left rhythmic
Bilateral rhythmic
Ictal EEG findings
31. Conclusions
• Ictal ESI is feasible in more seizures due to improved
algorithms
• Potential of ictal ESI is shown in several studies, but
most studies with small cohort
• Need for a large-scale prospective study.
• The added value of Ictal ESI in surgical decision
making should be studied compared to the added value
of other imaging techniques
ESI has been used primarily on interictal events because they are characterized by a higher signal-to-noise ratio (SNR) and are less contaminated by muscular artifacts compared to ictal events, and they are more likely to be captured during a recording session of limited duration (<1 h).
Summary: Source localization of epileptic foci using ictal spatiotemporal dipole
modeling (ISDM) yields reliable anatomic information in presurgical candidates. It
requires substantial resources from EEG and neuroimaging laboratories. The profile
and number of patients who may benefit from it are currently unknown. The purpose
of this study is to demonstrate the clinical usefulness of source localization in a
prospectively analyzed series. One hundred patients (51 male and 49 female patients)
with mean age of 31 years (range, 2 to 63 years) and mean duration of refractory
epilepsy of 20 years (range, 1 to 49 years) were enrolled consecutively in a presurgical
protocol. Ictal EEG was available in 93 patients. ISDM was performed when suitable
ictal EEG files were available. The clinical applicability of ISDM was examined in
three patients groups: 37 patients in whom ictal EEG recording and MRI were
congruent (group I), 30 patients in whom results were not completely congruent but not
incongruent (group II), and 26 patients in whom the results were incongruent (group
III). ISDM could be performed in 31 of 100 patients: 11 in group I, 8 in group II, and
12 in group III. ISDM influenced decision making in none of the patients in group I,
in 4 of 8 patients in group II, and in 10 of 12 patients in group III. Typically, the results
of ISDM directed avoiding intracranial EEG recordings in what appeared to be
unsuitable candidates for resection by clearly confirming the incongruency between
ictal EEG and MRI findings. In this series of 100 presurgical candidates, ictal source
localization could be performed in 31% of patients. In 14% of patients, it proved to be
a key element in the surgical decision process. Key Words: Electroencephalography—
EEG—Source localization—Dipole—Epilepsy surgery.
FIG. 1. Case no. 1: Patient no. 27, group
II, dipole type 1. A 37-year old righthanded
woman with a history of 28 years
of partial seizures. Video-EEG: bilateral
independent frontotemporal spikes; four
complex partial seizures bilateral recruitment.
Optimum MRI: left hippocampal atrophy
and T2 signal increase. FDGpositron
emission tomography: widespread
left hemisphere hypometabolism. Neuropsychology:
no focal functional deficits.
Ictal spatiotemporal dipole modeling: leftsided
type 1 dipole in all seizures. Surgical
decision: proceed with focal resection
(modified left-side temporal lobectomy
and hippocampectomy) without further invasive
recording. The patient has been seizure
free for more than 1 year.
FIG. 3. Case no. 3: Patient no. 36, group
III, dipole type 1. A 45-year-old righthanded
woman with a history of 34 years
of partial seizures. Video-EEG: bilateral
frontotemporal spikes; three complex
partial seizures: right frontotemporal recruitment.
Optimum MRI: left hippocampal
atrophy. FDG-positron emission
tomography: right temporal
hypometabolism. Neuropsychology: no
focal functional deficits. Ictal spatiotemporal
dipole modeling: right-side type 1
dipole in all seizures. Surgical decision:
not to proceed with invasive recording;
drop out of presurgical evaluation.
Selection and preparation of ictal EEG signals for the source analysis. Upper row, to left: EEG recording at the electrographic seizure start
(patient 12). Scale: 1 s 9 100 lV. FFT analysis is performed on successive segments of 1 s, with a 50% sliding window (upper row, in the
middle) until the decrease in peak frequency changes by 1 Hz (from 9.5 to 8.5 Hz). This way, the EEG epoch to be analyzed further is
delimited (red box within the EEG recording). Voltage maps are drawn at the negative peaks (lower row). Signals with similar voltage distribution
are averaged; in this figure the last voltage distribution (lower row, to the right) is different; therefore, it is not included into the
averaging. The averaged waveform is shown in the upper row to the right.
Ictal source localization in a patient with right mesial temporal focus (patient 15). (A) The averaged waveform. (B) Sequential voltage
maps on the ascending slope of the averaged waveform (timeframes: 18–26; duration of a timeframe: 4 msec). In each voltage map the
negative and positive peaks are marked automatically. A change in the voltage distribution along the ascending slope is observed. Source
localization is performed at timeframe 20 (blue box in Fig. 4B corresponding to the blue cursor line in Fig. 4A) and at timeframe 26 (green
box in Fig. 4B corresponding to the green vertical cursor in Fig. 4A). (C) The source localization corresponding to the initial voltage distribution
(timeframe 20) shows activation at the anterior part of the right temporal lobe/pole. (D) At the peak of the averaged waveform
(timeframe 26) the activation propagates to the posterior-lateral part of the temporal lobe and to the parietal lobe. The original ictal EEG
is showed in Data S5.
Figure 2.
(A) Illustrative result, patient 2: The EEG obtained with a 0.5–70 Hz band-pass filter and displayed in a bipolar montage. The first 2 s from
the seizure-onset (marker) was used for the analysis. (B) In the 256-electrode EEG the number of channels was reduced to 204. The dominant
frequency of the seizure pattern in this patient was 6 Hz, so the EEG was filtered with a bandwidth of 5–7 Hz. EEG displayed in a
monopolar montage. (C) Ictal ESI source of the 5–7 Hz frequencies projected on the individual MRI before surgery and (D) after surgery.
Figure 1.
Flowchart of the study. Positive reference: Patients with Engel class I for >1 year after surgery. Negative reference, patients with Engel
class III/IV for >1 year after surgery; Concordant localization, match between the ictal source localization and the reference standard;
Discordant localization, no match between the ictal source localization and the reference standard; Inconclusive, interictal ESI provided
two solutions due to the existence of two discrete equally dominant types of spikes.
Figure 3.
In patients 1, 2, 7, and 8, ictal ESI max (red circle) localized within the resected zone. Patient 5 had a first surgery in the left frontal lobe without seizure reduction and a second surgery of the left temporal lobe (where ESI max is localized) that led to seizure freedom. (B) Patient 3 is the only patient with a 128- electrode recording. Ictal ESI-max is localized 7 mm from surgical resection boundary. Patient 4 had a left occipital lobe resection, whereas ictal ESI localized seizure onset left temporal. Ictal ESI max for patient 6 is shown in the circle, whereas the patient underwent a right frontal lobe resection.
In 9 of 14 patients, interictal and ictal ESI solutions were concordant (patients 1, 2, 3, 7, 10, 11, 12, 13, and 14) and in agreement with the MRI findings in those who had an MRI lesion (patients 1, 2, 7, 13, and 14) (Table 1). In the MRI-negative cases (3, 10, 11, and 12), interictal and ictal ESI were also concordant. In patients 4, 5, 8, and 9, the interictal analysis provided more than one solution due to the existence of more than one equally prevalent type of spikes. It is notable that the ictal ESI was concordant with one of the interictal sources in each of these cases. In one case (patient 6), the interictal and ictal ESI were not concordant. In this case, interictal ESI localized the irritative zone in the right frontal pole, whereas ictal ESI-max was in the right frontocentral area.
PWA: Findet den stärksten Rhythmus zu jedem Zeitpunkt des Anfall und die räumliche Verteilung über die Elektroden. Ich glaub das kann man sich noch vorstellen. Der Rhythmus ist auf FT10 mit Stärke 1 auf F8 mit Stärke 0.8 usw.
Der Beamformer wandelt diese Elektrodeninfo des Rhythmus in 3D Info um, dazu wird ein Kopfmodell benötigt
The core idea of our ictal source localization (ISL) technique was to automatically determine the most dominant rhythmic EEG pattern within the earliest ictal activity, i.e. first change in EEG time-frequency plots. Next, we implemented a frequency dependent time window which had to contain at least eight ictal waves or discharges (e.g. 4 Hz ictal activity = time window of 2 seconds; 8 Hz ictal activity = time window of 1 second) to the selected ictal activity. The spatial distribution of this rhythmic activity over all EEG electrodes was the basis for our source localization method, leading to an automatic and artifact-robust localization approach. The inverse method used in our study was a frequency domain version of the minimum variance beamformer (MVB) based on a standard head model (Colin 27 Average Brain, Montreal Neurological Institute)32. The MVB tends to determine ictal activity as a more focal solution rather than a distributed one32; 33.
We localized seizures with high SNR without averaging and excluded seizures with very low SNR in a first step. Furthermore we used frequency domain techniques implemented in the minimum variance beamformer allowing for a simple suppression of noise and artifacts by focusing only on frequency parts belonging to the desired ictal signal and thus inherently increasing SNR32. Second, fast propagation of ictal activity in scalp-EEG may cause false localization. We aimed to localize the most dominant rhythmic EEG pattern within the earliest ictal activity, i.e. the first change seen in EEG time-frequency plots, and used a frequency dependent time window.
Purpose
To test the diagnostic accuracy of a new automatic algorithm for ictal source localization (ISL) during routine presurgical epilepsy evaluation following STARD (Standards for Reporting of Diagnostic Accuracy) criteria.
Methods
We included 28 consecutive patients with refractory focal epilepsy who underwent resective epilepsy surgery. Ictal EEG patterns were analyzed with a novel automatic ISL algorithm using a frequency domain version of the minimum variance beamformer based on a standard head model. ISL source localizations on a sublobar level were validated by comparison with actual resection sites and seizure free outcome 2 years after surgery.
Key Findings
Sensitivity of ISL was 92.3% and specificity 60%. Positive predictive value was 66.7%, negative predictive value 90%. The likelihood ratio was more than ten times higher for concordant ISL results as compared to discordant results. These performance measures were statistical significant (p=0.013).
Significance
Our ISL method may contribute to a correct localization of the seizure onset zone on a sublobar level and can readily be used in a standard epilepsy monitoring setting.