User:Priti Ram/Intracranial Electroencephalography

^[1]Intracranial electroencephalography is used to detect the candidates undergone extratemporal epilepsy surgery. For the studies some patients with normal MR findings and some with lesion were analysed by intracranial electrode assessments. Mayo Rochester Epilepsy surgery Database was used to detect the normal demographics of the patients. Patients with nonlesional extratemporal epilepsy can be determined by a focal high frequency (>20) oscillation at the onset of seizure in the iEEG and they also have Engel class I outcome after epilepsy surgery.Partial epilepsy is the common intractable epilepsy and the partial seizure is difficult to locate.Treatment for such epilepsy is limited to attachment of vagus nerve stimulator. Epilepsy surgery is the cure for partial epilepsy provided that the brain region generating seizure is carefully and accurately removed. Studies of neuronal activity in human and animals has shown a wide dynamic range of EEG activity. High frequency oscillations are related to onset of seizure in epileptic foci. On the other hand studies performed on human iEEG have limited dynamic range (~`0.1 to 40 Hz ) of iEEG activity. These limited frequencies are due to the fact that the oscillations with frequencies higher than 30 Hz are of low amplitude and are uncertain by cerebral activity at lower frequency range . So to detect high frequency oscillation at the onset of neocortical seizures broad band digital EEG recordings and spectral analysis can be used. HFEOS are the electrophysiology signature of epileptic brain and can be the reason for neocortical seizures. Although the above findings for refractory epilepsy is achieved, there is no clue for clinical significance for the same. Despite the fact that high frequency oscillations are the reason for onset of seizure in neocortical epilepsy, the data was extracted from the experiment containing small group of people. The data for improved localization of epileptogenic brain by interictal and ictal high frequency oscillation is also not known properly. An experiment was carried out to determine that HFEOs are the seizure onset frequency > 20 Hz and are the reason for achieving good surgical outcome in patients with normal MR findings undergoing Epilepsy surgery. Some patients with normal MR findings were selected for the experiment and underwent iEEG monitoring for presurgical evaluations of uncontrolled extra temporal epilepsy. For the process intracranial electrodes was placed according to presurgical standard protocols. Patients comprehensive analysis was performed using MRI, neuropsycological studies, visual perimetry and long term video EEG monitoring. Ictal and interictal spect and in some patients SISCOM was performed. Subdural grids, strips and depth electrodes were used.

IEEG data collection and storage was done by XLTEK or NCI. High-pass filter of 1.0 HZ and low-pass filter of 100 Hz, bipolar montage electrodes were used. To eliminate line noise a digital 60 Hz notch was used.

Determination of seizure and seizure onset zone

The earliest change in the EEG from the baseline was selected as the seizure point in the neocortical region. The occipital, frontal, Parietal zone was determined with the electrodes showing earlist seizure onset activity. The seizure onset pattern of iEEG was characterised by waveform morphology, frequency of discharge,spatial extent. Arrangements of subdural grids with patients postresection MR images were used evaluate the previously mentioned EEG features and to locate epileptogenic region. The epileptogenic region when resected produce seizure control. The arrangements of subdural grids and post resection MR image allowed to detailed correlation of ictal and interictal EEG measurements used to localize epileptogenic region (size,focal) and also anatomical structures (frontal, parietal and occipital lobes). Anatomical structures allowed complete resection of epileptogenic brain.

Neocortical seizure was determined when seizure onset preceded by Greater than or equal to a minute uninterrupted free interictal EEG. Data analysis that is seizure onset times and seizure onset zone was done by visual review and spectral analysis (MATLAB,Inc , 0.25 Secs sliding window and 0.2 secs overlap). The frequency of oscillation at seizure was determined by selecting the dominant spectral peak. Due to limited number of patients, seizures with onset frequencies more than 20 Hz were grouped under HFEO.

The connection between seizure pattern, spect location and seizure outcome were determined by Chi- Square and Fischer test when there were less than 5 patients in each group. An association between total of electrodes and seizure outcome was tested by rank sum statistics. The probability value of greater than equal to 0.05 was used as statistically significant.

Results: 31/51 patients undergoing iEEG underwent epilepsy surgery. In 20 patients adequate information of seizure origin localization was not obtained and hence these patients were not considered suitable for resection. The location of grid location varied in each patients and obtained by seizure signals, results of scalp EEG, spect and SISCOM studies. 21/31 patients who underwent epilepsy surgery information regarding follow-ups of more than a year, seizure frequency and iEEG recordings was available. The average age at seizure onset was 8.5 yrs and average duration of seizures was 16.5 yrs.

The number of subdural electrodes varied in different patients from  24 to 128 electrodes and reflecting the degree of localization based on non-invasive presurgical evaluation. However number of subdural electrodes did not correlate with seizure outcome from surgery.

All patients in this study underwent ictal and interictal spect analysis. In 15/31 patients who underwent resection, region of spect abnormality was identified. However, there was no correlation between spect abnormality and seizure- free outcome.

The most common iEEG pattern at seizure onset in the surgical treated group was a focal HFEO (15/28), 3 patient had diffuse HFEO , 4 patients had rhytmic spiking or gama wave activity, 2 had alpha-teta activity. 26 patients had frontal lobe neocortical and 2 had pareital lobe resection. In 25/28 patients who underwent resection gliosis was noted and remainder had cortical dysplasia.

There was no significant relation between seizure type and region where seizure occured. Patient with cortical dysplasia had high frequency seizure onset. 10/28 surgically treated patients were seizure free. Fourteen had Engel class 1 outcome, seventeen patients had 80% seizure reduction. A focal HFEO at seizure onset was found in 14 patients who had class 1 Engel outcome and was unrelated in patients with Engel class 2-4 outcome.  Patients with difuse high frequency onset or gamma wave activity at onset did not had  good surgical outcome. No pathological correlation were available for seizure location and seizure outcome.

Discussions

Seizure outcomes are favourable mostly in patients with mesial temporal lobe epilepsy and lesional neocortical lesion. Consistent results of electrophysiological studies MR imaging had a predicative value for excellent surgical outcomes. However MR imaging is ineffective in localizing lesion  in 20-30% patients especially those with cortical dysplasia and Gliosis