180
K.-Y. Jung
beyond the area of resection correlates with poor surgical outcome in patients with
extrahippocampal epilepsy [3]. Thus, identification of interictal spike area is helpful
for surgical treatment of epilepsy.
There are several studies reporting various levels of accuracies of interictal ESI.
Inconsistent results may be partly attributed to the complicated steps for ESI and
absence of standardized methods for ESI. The processes for ESI include generation
of head model from brain MRI, calculation of inverse solution, and co-registration of
electrode coordinates with brain MRI (refer to Fig. 1 in Brodbeck et al. [8]). There are
various methodologies that can be possibly adopted for ESI, in terms of brain MRI
(individual vs. standard template), head model (spherical, boundary element method,
and finite element method), number of electrodes, electrode coordinates (individual
vs. standard), and inverse methods (discrete vs. distributed). As numerous methods
can be generated from the combination of the above-mentioned methods at each step
of ESI, the diverse results are inevitable.
Furthermore, a variety of gold standards for estimating accuracy of ESI have
been adopted, such as phantom simulation, comparison with spike location using
intracranial electrodes, distance from surgical resection margin, and postoperative
surgical outcome. For these reasons, care should be taken not to misinterpret the
results of ESI. The best way to assess clinical impact of ESI on epilepsy surgery
seems to evaluate the correlation of the results of ESI with prospective long-term (a
minimum 2 years follow-up) postoperative surgical outcomes. There are only few
studies referring to surgical outcomes.
Brodbeck and his colleagues prospectively evaluated clinical utility of ESI as a
part of the pre-surgical work-up from 152 patients with a variety of partial epilepsies
who underwent surgical treatment for intractable seizures [8]. Two third of patients
had temporal lobe epilepsy and 77% had good surgical outcome, which was assessed
at least 1 year after surgery. Standard clinical EEG using less than 32 electrodes was
obtained basically, and a high-resolution EEG with 128 or 256 electrodes was also
recorded in 55 patients. However, the method for coordination of electrode location
(i.e., measured vs. standard template coordinate) was not described. A simple realistic
head model using spherical model with anatomical constraints (SMAC) method was
generated from individual MRI, and the linear distributed inverse algorithm known
as local autoregressive average (LAURA) was used to estimate intracranial sources
for IEDs. The rising phase of averaged IED from the most prevalent IEDs was subject
to the source estimation.
Good surgical outcome (Engel class I and II) was considered to be the gold standard for correct localization of the epileptogenic focus. Sensitivity was defined as
the percentage of patients with focus localization within the resected zone of all
patients who were seizure-free (n 117). Specificity was defined as the percentage
of patients with localized focus outside the resected zone of all patients who had
an Engel Class III or IV outcome after surgery (n 19). Sensitivity and specificity
were compared among various source estimation conditions, and among other neuroimaging methods. The sensitivity and specificity of ESI were highest when high
resolution EEG and individual MRI were adopted, which showed 84.1 and 87.5%,
respectively. When compared to other imaging modalities, these values were higher
K.-Y. Jung
beyond the area of resection correlates with poor surgical outcome in patients with
extrahippocampal epilepsy [3]. Thus, identification of interictal spike area is helpful
for surgical treatment of epilepsy.
There are several studies reporting various levels of accuracies of interictal ESI.
Inconsistent results may be partly attributed to the complicated steps for ESI and
absence of standardized methods for ESI. The processes for ESI include generation
of head model from brain MRI, calculation of inverse solution, and co-registration of
electrode coordinates with brain MRI (refer to Fig. 1 in Brodbeck et al. [8]). There are
various methodologies that can be possibly adopted for ESI, in terms of brain MRI
(individual vs. standard template), head model (spherical, boundary element method,
and finite element method), number of electrodes, electrode coordinates (individual
vs. standard), and inverse methods (discrete vs. distributed). As numerous methods
can be generated from the combination of the above-mentioned methods at each step
of ESI, the diverse results are inevitable.
Furthermore, a variety of gold standards for estimating accuracy of ESI have
been adopted, such as phantom simulation, comparison with spike location using
intracranial electrodes, distance from surgical resection margin, and postoperative
surgical outcome. For these reasons, care should be taken not to misinterpret the
results of ESI. The best way to assess clinical impact of ESI on epilepsy surgery
seems to evaluate the correlation of the results of ESI with prospective long-term (a
minimum 2 years follow-up) postoperative surgical outcomes. There are only few
studies referring to surgical outcomes.
Brodbeck and his colleagues prospectively evaluated clinical utility of ESI as a
part of the pre-surgical work-up from 152 patients with a variety of partial epilepsies
who underwent surgical treatment for intractable seizures [8]. Two third of patients
had temporal lobe epilepsy and 77% had good surgical outcome, which was assessed
at least 1 year after surgery. Standard clinical EEG using less than 32 electrodes was
obtained basically, and a high-resolution EEG with 128 or 256 electrodes was also
recorded in 55 patients. However, the method for coordination of electrode location
(i.e., measured vs. standard template coordinate) was not described. A simple realistic
head model using spherical model with anatomical constraints (SMAC) method was
generated from individual MRI, and the linear distributed inverse algorithm known
as local autoregressive average (LAURA) was used to estimate intracranial sources
for IEDs. The rising phase of averaged IED from the most prevalent IEDs was subject
to the source estimation.
Good surgical outcome (Engel class I and II) was considered to be the gold standard for correct localization of the epileptogenic focus. Sensitivity was defined as
the percentage of patients with focus localization within the resected zone of all
patients who were seizure-free (n 117). Specificity was defined as the percentage
of patients with localized focus outside the resected zone of all patients who had
an Engel Class III or IV outcome after surgery (n 19). Sensitivity and specificity
were compared among various source estimation conditions, and among other neuroimaging methods. The sensitivity and specificity of ESI were highest when high
resolution EEG and individual MRI were adopted, which showed 84.1 and 87.5%,
respectively. When compared to other imaging modalities, these values were higher