ECOG

1. ELECTROCORTICOGRAPHY
(ECOG)
Lalit Bansal, M.D.

3. NEED for ECOG
• Scalp EEG is limited - activity from deep or mesial brain areas are not recorded
• Discharges recorded in the scalp are attenuated and distorted by different layers (scalp, bone, meninges)
The skull attenuates the signal considerably (approx. factor 5)
Example:
 Generator in the hippocampus, potential of 500 μV, 0.5 cm from the generator
 4 cm from the generator (8 times further), on the temporal neocortex, the potential is 64 times smaller(8 μV)
 On the scalp, the potential is reduced by a factor of 5 (1.6 μV)
• 6cm2 of brain cortex is needed to discharge synchronously to generate detectable potential in scalp EEG.
Activity generated by a smaller cortical surface, substandard amplitudes, or midline structures are not
reflected on scalp EEG
• Difficult to localize and determine the extend of the cortical area from scalp EEG
• The skull attenuates frequencies above 10 000 Hz

4. Intraoperative ECOG
• Initially pioneered in human patients by Hans Berger 1929
• First used by Penfield and Jasper in 1950 to map epileptic activity
• Electrodes can be placed almost anywhere with a relatively low risk
• ECOG is of very high quality, without the main artifacts visible on the scalp (movement, eye movements,
EMG)
• Electrodes can be very close to neuronal generators
• Electrodes can be used for stimulation, to trigger seizures or for functional mapping

5. Types of ECOG:
1. Intraoperative ECOG
2. Extraoperative ECOG

7. Indications for combination of ECoG and functional mapping
I General indication:
Whenever the irritative zones/epileptic foci are or could be in close proximity to eloquent cortex
II Goals Guiding the surgical approach
1. Extent of resection 2. Subpial transections
III Specific situations
1. Left temporal foci 2. Frontal or parietal foci
3. Multifocal epilepsy 4. Neurodevelopmental lesions

8. Electrodes
• Recording electrodes usually made of Platinum-iridium or stainless steel or less commonly carbon or silver
• Platinum slightly more stable when current is passed and used with MRI/MEG
• Stainless steel may possibly diffuse metal ions across the electrode-pial interface (significant over longer
periods of stimulation)
Montage
• Bipolar
• Referential - identify complex electrical spike field
- avoid bipolar cancellation of potentials
Voltage
• 5-6 times higher than scalp EEG
• around 300-500 microV
• sensitivity often set at 30-70 microV/mm
ECOG duration
• Minutes (intraoperative) to days (extraoperative 7-14 days)

9. Recording Technique
Sampling:
• Usually at 500 Hz but with increased recognized relevance of HFO, sampling rate of 1000-2000 Hz is used (prevent
aliasing)
Amplification:
• ECOG signal need amplification and amplifier must have – a) amplitude linearity, b) adequate bandwidth, c) phase
linearity, d) low noise
Filtering:
• Depending of waveform that need to be seen (low freq vs HFO)
Electrical safety:
(1) using power receptacles and adaptors with a ground prong
(2) connecting all the patient associated equipment to 1 cluster of power receptacles, ideally with separation between
lines and with removal of other sources of power from the patient environment
(3) avoiding contact between the patient and low-resistant pathways such as metal beds, plumbing, metal architectural
elements, or liquids

10. Influence of antiepileptic drug reduction or withdrawal
• Absent or subtherapeutic drug level leads to increase seizure frequency
• Secondary generalization is more often seen after drug reduction/withdrawal
• Interestingly – localization, morphology at seizure onset, time to spread to contralateral and coherence of EEG
discharges do not change
• During seizure cluster, some areas of brain can start to seize that usually do not spontaneously
Influence of Anesthetic agents
• Enhance epileptiform activity – methohexital, fentanyl, remifentanil, propofol, thiopental
• Decreases or increases epileptiform activity – isoflurane, sevoflurane
• Dexmedetomidine (Precedex) – alpha 2 receptor agonist – produce natural sleep pattern, reduce need for propofol
Factors modifying ECOG/epileptiform activity

11. Electrical stimulation of the cortex
• Electrical stimulation of cortex can elicit after discharges, subclinical seizures, habitual or nonhabitual auras, and
habitual and nonhabitual clinical seizures
• Electrically induced auras and seizures frequently correlate with seizure onset zone (unilateral focus) (concordance
75-100% , Schulz et al., 1997; Bernier et al., 1990)
• >1 Seizure onset focus - have decrease concordance
• After-discharge threshold not a reliable predictors of spontaneous SOZ
• After discharge, after discharge evolving to clinical seizure, or after discharge > 10 seconds – does not
topographically correlate to SOZ (Blume et al., 2004)
• Localization value of auras and clinical seizures is superior to after discharges

12. Volume (invasiveness) of depth electrodes
• Depth contact 50 mm length:
 If 0.8 mm diameter, volume 25 mm3
 If 1.2 mm diameter, volume 56 mm3
• If we place 20 electrodes = 500 to 1000 mm3 or 0.5 to 1 cm3
• Brain volume 1300 cm3
• Sampling area = 0.5 /1300 = 0.04%
• Sampling area = 1/1300 = 0.08%
One electrode explores ~5mm brain volume beyond its border
• 1 multi-contact electrode «captures» a cylinder of ~ 1 cm in diameter and 6 cm in length= 5 cm3
• 10 electrodes (100 contacts) = 50 cm3
• One hemisphere ~ 650 cm3
• 10 electrodes = 8% of one hemisphere

13. Subdural grids
• 1 grid contact captures a disk of ~ 2 cm in diameter (3 cm2)
• One 8 x 8 grid explores 64 x 3 ~ 200 cm2
• One hemisphere~ 5,000 cm2
• One 8 x 8 grid covers about 4% of the cortical surface of one hemisphere

14. Different generator size, same spike amplitude
• The amplitude of the scalp EEG is largely function of the size (surface) of the generator. EEG
electrodes are at approximately the same electrical distance to the generator
• The amplitude of the ECOG is largely a function of the distance between the electrode and the
generator

15. 1. Repetitive electrographic seizures
• Recruiting/derecruiting frequency around 12 to 16 Hz.
2. Repetitive bursting patterns
• High frequency (10 to 20 Hz) lasting for 5 to 10 seconds
3. Continuous or quasi-continuous rhythmic spiking
• Prolonged trains of rhythmic 2-8 Hz spikes
ECoG Patterns Associated with Cortical Dysplasia
Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37:476-87

16. Recruiting/derecruiting epiteptogenic pattern, characterizing an electrographic seizure,
maximum over the inferior occipital region

17. Repetitive bursts of polyspikes, with variable amplitude and duration, recorded
diffusely from the right fronto-central regions

18. Hippocampus--Lateral Neocortex

19. Continuous, rhythmic, or semirhythmic spikes, recorded from centro-temporo-parietal
electrode

20. (1) Sporadic spikes: spikes occurring at irregular
time intervals at several sites
(2) Continuous spiking: spikes occurring
rhythmically at regular time intervals for at least
10 seconds, the interval between two subsequent
spikes being 1 second at the most (frequency ≥ 1
Hz)
(3) Bursts of spikes: sudden occurrence of spikes
for at least 1 second with a frequency of 10 Hz or
greater
(4) Recruiting discharges: rhythmic spike activity
characterized by increased amplitude and
decreased frequency
Electrocorticographic Patterns in Epilepsy Surgery
and Long-Term Outcome
San-Juan, et al. 2017

21. Reliability of Intraoperative ECOG

22. • Intraoperative ECOG in children in intractable neocortical epilepsy reliable if spike frequency > 10 spike/min1
• Spike frequency <1/min is largely unreliable for localization of seizure focus1
• In temporal lobe epilepsy, pre-resection ECOG showing >18spike/min associated with good outcome2 and
<1spike/4 min associated with poor outcome
1Asano E, Benedek K, Shah A, et al. Is intraoperative electrocorticography reliable in children with intractable neocortical epilepsy. Epilepsia 2004;45(9):1091-9
2McBride MC, Binnie CD, Janota I, Polkey CE. Predictive value of intraoperative electrocorticograms in resective epilepsy surgery. Ann Neurol 1991;30:526-32
Reliability of Intraoperative ECOG

23. Post Surgical Interictal Spikes and Seizure Outcome

24. Post Surgical Interictal Spikes and Seizure Outcome
• Arise from unresected epileptogenic tissue with potential to cause seizures
• Secondary to cortical isolation, to surgical trauma or activation secondary to partial excision
• Surgical isolation of normal cortex can cause burst suppression, spike-burst suppression
• Surgical injury to cortex can cause postoperative spikes
• Activation of partial excision – minor secondary focus that was suppressed by its proximity to a dominant
focus

25. Post Surgical Interictal Spikes and Seizure Outcome
• Arise from unresected epileptogenic tissue with potential to cause seizures
• Secondary to cortical isolation, to surgical trauma or activation secondary to partial excision
• Surgical isolation of normal cortex can cause burst suppression, spike-burst suppression
• Surgical injury to cortex can cause postoperative spikes
• Activation of partial excision – minor secondary focus that was suppressed by its proximity to a dominant
focus
 No general agreement in the value of postoperative ECoG to predict seizure outcome (Nair and Najm, 2008)
 In 80 patients with refractory temporal epilepsy, the presence of spikes in postresection ECoG was more
frequent in patients with postoperative seizures (72%) than in seizure-free patients (47%) (Fiol et al., 1991)
 In 87 patients with temporal lobe epilepsy, postresection spikes and the change from preresection to
postresection spikes did not correlate with seizure outcome (Kanazawa et al., 1996)

26. Post Surgical Interictal Spikes and Seizure Outcome
• In 47 patients with medial temporal lobe epilepsy, there was no correlation between the presence or absence of
postresection spikes and seizure outcome (Tran et al., 1995)
• In 36 patients with brain tumor and seizures who underwent resection limited to the tumor margins, a correlation
between preresection or postresection spikes, and seizure outcome could not be found (Tran et al., 1997)
• In 94 patients with refractory temporal epilepsy, the presence of spikes in the postsurgical ECoG did not predict
seizure outcome (Benifla et al., 2006)
• In small study of 15 patients with refractory temporal lobe epilepsy, there was nonsignificant better seizure outcome
in patients with spikes than in patients without spikes in the postresection ECoG: seizure freedom was achieved by 8
of 10 (80%) patients with residual spikes and by 3 of 5 (60%) patients with no postsurgical spikes (Chen et al., 2006)
• In 140 patients with refractory mesial temporal lobe epilepsy - patients with hippocampal (but not cortical or
parahippocampal) spikes in postresection ECoG had a significantly worse seizure outcome (McKhann et al., 2000)
 Engel class I seizure outcome was achieved in 29% patients with hippocampal spikes
 73% patients who had no spikes in the hippocampus
 76% patients with no spikes at all
• Postresection ECoG may be of prognostic significance, especially when spikes are residual, not newly appearing spikes
(Binnie et al., 2001)

27. Functional Mapping and Cortical Stimulation
• Luigi Rolando first used galvanic current to stimulate cerebral cortex of living animals in 1809
• Fritz and Hitzig in 1870 first used cortical stimulation in dogs to delineate motor cortex
• Performed to optimize the extend of resection and minimize or avoid deficit
• Bipolar stimulation if used for FM with cathode/anode at the target tissue
• Potential brain injury can happen with cortical stimulation and is depended on charge density
• Charge density is the function of charge and cross-sectional area of the electrode surface in contact with the
brain
• Brain injury is usually seen with continuous stimulation. Never been reported with intermittent stimulation
(which is commonly used in humans for FM)

28. • To avoid tissue damage, critical measurement is charge density measured in μCoulombs/cm2/phase
• Safe maximum is considered to be 50 to 60 μC/cm2/phase
• Charge (Coulombs) = current density (I) x pulse duration (D)
• 1 μC = 1 mA x 1 ms = 2 mA x 0.5 ms
• Charge Density CD = C/ACE (ACE = area of stimulating electrode)
• Subdural grid leads will deliver charge densities of 54 to 57 microcoulombs/cm2 per phase for peak
currents of 13.6 to 15 mA
Charge Density

29. Stimulation intensity (mA) plotted against charge density ([mu]C/cm2) of commonly used SEEG, grid/strip, and
intraoperative probe electrodes (top). Charge density is segregated into "safe," "risky," and "dangerous" categories
based on criteria used by the FDA for the approval of the predicate stimulator device (bottom). These safety
criteria do not take into account other important factors in safety, including interelectrode distance, and presence
(electrocorticographic) or absence (stereo-electroencephalographic) of current shunting through cerebrospinal
fluid, for example.
Animal studies with continuous
stimulation of upto 50 hours

30. Geometry (to scale) of exposed "effective" surfaces available for stimulation of commonly used surface
(grid/strip), depth (stereo-electroencephalographic) and probe (intraoperative handheld) electrodes

31. Oscillatory classes in the cortex
Ripple: 80-150 Hz
Fast Ripple: 150 – 500 Hz
Ultra Fast Ripple: 500 – 2000 Hz

32. Ripple classification helps to localize the seizure‐onset zone in neocortical epilepsy
Epilepsia, Volume: 54, Issue: 2, Pages: 370-376, First published: 25 October 2012, DOI: (10.1111/j.1528-1167.2012.03721.x)
Pathological ?Physiological
(w/o spike)

33. The 14&6/sec positive spikes normal EEG variant is correlated exclusively with hippocampal activity
(Kokkinos. 2019)
The variant is time-locked to high-amplitude spike bursts overlaid on low-amplitude slow waves

36. Seizure Onset from anterior hippocampus with early spread to amygdala

38. Seizures from end chain electrodes not reliable for seizure localization
Ictal onset and spread can look similar

39. 2 x 6 strip lateral L temp, 1 anterior
1 x6 strip inferior L temp, 1 mesial
1 x 4 strip L temporal pole, 1 wrapped at pole

40. 2 x 6 strip lateral L temp, 1 anterior
1 x6 strip inferior L temp, 1 mesial
1 x 4 strip L temporal pole, 1 wrapped at pole
Attenuated anterior L temporal with burst suppression (C1-3, C7-9)
Pulse artifact C13-14, C17-18
Attenuated anterior lateral L temporal at pole, asynchronous spike waves

42. Pulse Artifact on ECOG

43. 4x5 grid – sylvian fissure – tumor case

44. 4x5 grid – sylvian fissure – tumor case
Anesthesia induced fast activity

45. LFF 80 Hz; HFF – OFF; Notch- OFFLFF – 1 Hz; HFF- 70 Hz; Notch - ON

46. LFF -1 HZ; HFF – 70 Hz; Notch ON

47. LFF- 100Hz; HFF – OFF, Notch OFF

48. Thank You !
Questions?