8 Analysis of EEG in Medically Intractable Epilepsy
183
Since the aim of resective epilepsy surgery is to achieve freedom of seizure by complete elimination of epileptogenic zone, ictal EEG is considered to be more important
than IEDs to determine the resection margin in surgical treatment of epilepsy. However, ictal ESI is largely limited by several reasons. For example, in a series of 100
presurgical candidates, ictal dipole source localization could be performed in only
31% of patients [5]. One of the reasons is that seizures with vigorous motor activities
inevitably accompany many muscle and movement artifacts which lower signal-tonoise ratio hampering proper source localization. Ictal EEG activities often show
bilateral or nonlateralized pattern, which also makes ESI difficult to correctly localize SOZ. It is not uncommon that patient do not have seizure during long-term EEG
monitoring. Another important reason is that ictal rhythm can only be seen at the
scalp when the cortical areas are sufficiently synchronized. Furthermore, seizures
originating from medial side of cortex or deep sulci cannot be picked up by scalp
electrodes until ictal activities propagate to lateral or superficial cortices. This means
that ictal ESI identifies not only the ictal onset zone but also the cortex to which
seizure discharges spread during an early ictal event [22].
Contrary to interictal ESI, there are relatively few clinical studies on the source
localization of the ictal epileptiform activity. Source localization of rhythmic ictal
scalp EEG activity has been reported on 42 patients with focal epilepsy, aged between
9 and 69 years (mean 35.9, median 38 years) by Danish group [4]. EEG was recorded
using a standard 25 electrode (19 electrodes of the international 10–20 system with
extra-electrodes in the inferior temporal chain). For the sake of clinical feasibility,
standard electrode positions and template MRI (MNI template) were used for source
localization in this study. The earliest rhythmic ictal activities that are stable in terms
of frequency identified by spectral analysis and spatial distribution identified by
voltage map were selected for the ictal ESI. The duration of these epochs ranged
between 1 and 11 s (median 3 s). LAURA was used for the source localization. The
predictive values were estimated based on the surgical outcome, as evaluated 1 year
after the operation.
Ictal source localization at the sublobar level could be performed in 72% of the
patients who had at least one epileptic seizure in the presurgical evaluation. The
sensitivity of the ictal ESI was 69.7% and the specificity was 75.7% with regard to
reference standard that was determined by the consensus conclusion of the multidisciplinary epilepsy surgery team. It has been reported that the seizure onset at the lobar
level with reference to 2-year surgical outcome was correctly localized in 76–83% of
temporal seizures and 47–65% of extratemporal seizures by visual inspection of scalp
ictal EEG [43]. However, only 21–38% of scalp ictal EEG agreed with depth EEG
at the sublobar level [38]. Therefore, it is remarkable that ictal ESI can significantly
increase the accuracy of localization with higher precision than visual interpretation
of scalp ictal EEG. Twenty patients underwent surgery and 16 patients (80%) became
seizure-free. The PPV of the ictal source localization was 92%, and the NPV was
42.8%. Furthermore, the positive likelihood ratio for the concordant results (ictal
source localization matching the reference standard) was 3.0 which is the nine times
higher than the negative likelihood ratio of the discordant results (0.33, mismatch
between ictal source localization and the reference standard).
183
Since the aim of resective epilepsy surgery is to achieve freedom of seizure by complete elimination of epileptogenic zone, ictal EEG is considered to be more important
than IEDs to determine the resection margin in surgical treatment of epilepsy. However, ictal ESI is largely limited by several reasons. For example, in a series of 100
presurgical candidates, ictal dipole source localization could be performed in only
31% of patients [5]. One of the reasons is that seizures with vigorous motor activities
inevitably accompany many muscle and movement artifacts which lower signal-tonoise ratio hampering proper source localization. Ictal EEG activities often show
bilateral or nonlateralized pattern, which also makes ESI difficult to correctly localize SOZ. It is not uncommon that patient do not have seizure during long-term EEG
monitoring. Another important reason is that ictal rhythm can only be seen at the
scalp when the cortical areas are sufficiently synchronized. Furthermore, seizures
originating from medial side of cortex or deep sulci cannot be picked up by scalp
electrodes until ictal activities propagate to lateral or superficial cortices. This means
that ictal ESI identifies not only the ictal onset zone but also the cortex to which
seizure discharges spread during an early ictal event [22].
Contrary to interictal ESI, there are relatively few clinical studies on the source
localization of the ictal epileptiform activity. Source localization of rhythmic ictal
scalp EEG activity has been reported on 42 patients with focal epilepsy, aged between
9 and 69 years (mean 35.9, median 38 years) by Danish group [4]. EEG was recorded
using a standard 25 electrode (19 electrodes of the international 10–20 system with
extra-electrodes in the inferior temporal chain). For the sake of clinical feasibility,
standard electrode positions and template MRI (MNI template) were used for source
localization in this study. The earliest rhythmic ictal activities that are stable in terms
of frequency identified by spectral analysis and spatial distribution identified by
voltage map were selected for the ictal ESI. The duration of these epochs ranged
between 1 and 11 s (median 3 s). LAURA was used for the source localization. The
predictive values were estimated based on the surgical outcome, as evaluated 1 year
after the operation.
Ictal source localization at the sublobar level could be performed in 72% of the
patients who had at least one epileptic seizure in the presurgical evaluation. The
sensitivity of the ictal ESI was 69.7% and the specificity was 75.7% with regard to
reference standard that was determined by the consensus conclusion of the multidisciplinary epilepsy surgery team. It has been reported that the seizure onset at the lobar
level with reference to 2-year surgical outcome was correctly localized in 76–83% of
temporal seizures and 47–65% of extratemporal seizures by visual inspection of scalp
ictal EEG [43]. However, only 21–38% of scalp ictal EEG agreed with depth EEG
at the sublobar level [38]. Therefore, it is remarkable that ictal ESI can significantly
increase the accuracy of localization with higher precision than visual interpretation
of scalp ictal EEG. Twenty patients underwent surgery and 16 patients (80%) became
seizure-free. The PPV of the ictal source localization was 92%, and the NPV was
42.8%. Furthermore, the positive likelihood ratio for the concordant results (ictal
source localization matching the reference standard) was 3.0 which is the nine times
higher than the negative likelihood ratio of the discordant results (0.33, mismatch
between ictal source localization and the reference standard).