Inter-hemisphere Asymmetries in EEG Signal Properties in Patients with Temporal Lobe Epilepsy versus Non-epileptic Seizures
Abstract number :
2.136
Submission category :
3. Clinical Neurophysiology
Year :
2011
Submission ID :
14872
Source :
www.aesnet.org
Presentation date :
12/2/2011 12:00:00 AM
Published date :
Oct 4, 2011, 07:57 AM
Authors :
H. J. Skinner, D. S. Shiau, J. J. Halford, J. C. Sackellares
Rationale: Non-epileptic seizures (NES) and epileptic seizures can appear clinically similar, which can lead to erroneous diagnosis and treatment. Differentiation of these two conditions often requires multi-day EEG monitoring to capture interictal or ictal epileptiform abnormalities. Also, if these epileptiform abnormalities are not seen during the EEG monitoring, then a diagnosis cannot be made. Therefore, it would be beneficial to develop a diagnostic method that can reliably distinguish EEG properties of patients with NES from those with epileptic seizure by analysis of brief epochs of EEG. Our goal in this preliminary study was to investigate whether differences in EEG signal measurements of inter-hemisphere asymmetry (IHA) exist between patients with NES and those with temporal lobe epilepsy (TLE).Methods: Interictal EEG epochs (at least 10 seconds each) were sampled from long-term epilepsy monitoring unit recordings obtained from 62 patients. A total of 620 epochs in the relaxed, awake state were sampled from TLE and NES patient groups. Within each EEG sample epoch, we calculated the signal regularity using the pattern-match regularity statistic (PMRS). Signal power was calculated using amplitude variation (AV). These calculations were performed in the F8, T4, F7, and T3 EEG channels utilizing a non-overlapping 5.12 second computation window. IHA values were then calculated as the difference between left (F7 and T3) and right (F8 and T4) channels, with respect to PMRS and AV values, respectively. Outlier PMRS and AV values were excluded using Grubb s test. PMRS variable were found to follow a normal distribution. Two-tailed, independent sample t-tests were applied to test for significant difference (p<0.05) between the two patient groups with regard to symmetry of the PMRS within the Temporal EEG channels. AV values did not follow a normal distribution. The Whitney-Mann-U test was used to test the hypothesis that inter-hemispheric AV differed between the two groups. Results: Of the 62 patients, 29 were NES patients. Of the 33 TLE patients, 14 had seizures with onset recorded exclusively from the left Temporal electrodes, 11 had only right Temporal seizure onset, and 8 had seizures that began in both the left and right Temporal electrodes. A total of 71 outliers were excluded (60 AV and 11 PMRS). No significant difference in AV IHA values was found between the TLE and NES groups (2.9261 vs. 2.6379, p=0.955). However, when comparing right-left symmetry of the PMRS, TLE patients had significantly higher values (0.39397 vs. 0.19426, p=0.004). Conclusions: In this preliminary study, we found that asymmetry of inter-hemisphere signal regularity from the Temporal electrodes is significantly larger in patients with TLE than those with NES. These results suggest it may be possible to develop an algorithm that will estimate the probability of a patient having epilepsy seizures rather than NES.
Neurophysiology