- Email: [email protected]
Predictive value of intraoperative electrocorticography in seizure outcome after temporal lobectomy
18P
Society Proceedings/Electroencephalography
16. EEG spectral findings in healthy, elderly men and women - sex differences. - R.P. Brenner qb, RF. Ulrich ’ and C.F. Reynolds, III a,’ (Departments of ’ Psychiatry and b Neurology, University of Pittsburgh, School of Medicine, Pittsburgh, PA) There are few quantitative EEG studies evaluating gender-related differences in healthy, elderly subjects. Therefore, we performed computerized spectral analysis of the EEG in 119 healthy, elderly men and women, ranging in age from 60 to 87 years with a mean age of 70.4 years, to evaluate gender-related variance. The groups did not differ in age, education levels or Folstein score. We compared men (n = 53) to women (n = 661, controlling for the effect of age by decade. Compared to men, women had a significant (usually P < 0.01) increase in the parasagittal mean frequency (l-30 Hz and 4-20 Hz), beta, and beta,, while alpha, and theta-beta were decreased. Although the mean frequency was lower in older subjects, there was no significant difference by decade. In addition, there were no significant sex-age interactions. Our study adds to the existing literature of adult spectral EEG gender-related differences and extends it to late-life subjects. Although the basis for these differences is uncertain, we feel that quantitative EEG must control for variance related to gender. 17. Spatiotemporal dynamic analysis of electroconvulsive seizures. - A.D. Krystal, H.S. Greenside ‘, R.D. Weiner Gassert (Department of Psychiatry, Duke University Center, and “Departments of Physics and Computer Duke University, Durham, NC)
therapy and D. Medical Science,
The physiology of electroconvulsive therapy (ECT) seizures is poorly understood. It has been assumed that these seizures are a state of decreased complexity in brain function (neuronal hypersynchrony) compared with waking. Yet, the spatiotemporal properties of the EEG recorded during ECT seizures (ictal EEG) have not been studied quantitatively, despite evidence that ictal EEG properties relate to ECT side effects and efficacy. We preliminarily studied the spatiotemporal dynamics of 21-channel ictal EEG data recorded from 9 depressed subjects clinically referred for ECT. Coherence was calculated for multiple pairs of electrodes to reflect the pattern of the relationship of brain electrical activity across the scalp. The local largest Lyapunov exponent (LI was calculated for each channel to reveal the spatial pattern of temporal complexity across the scalp. These indices were calculated for the waking EEG recorded before the ECT course, the post-anesthesia (but pre-stimulus) EEG, and the ictal EEG. Compared with the waking and post-anesthesia states, we found a marked increase in the coherence of EEG activity across the scalp and evidence for a marked decrease in both temporal and spatial complexity of the ictal EEG (decrease in L across the whole scalp) during the seizures. 18. Computer-assisted seizure detection based on quantitative dynamical measures. - J.C. Sackellares, L.D. Iasemidis, A. Berreto, R.L. Gilmore, R.A. Savit ‘, S.A. Roper and B.M. Uthman (Departments of Neurology, Neurosurgery and Electrical Engineering, University of Florida, Gainesville, FL, and * Department of Physics, University of Michigan, Ann Arbor, MI) Dynamical measures developed to study complex systems provide a quantitative description of the time-varying properties of the EEG. We have found characteristic changes in dynamical measures such as the Lyapunov exponent and time dependency indices during partial seizures of temporal lobe origin. The Lyapunov exponent reflects the chaoticity of the signal. The time dependency indices reflect the temporal structure of the signal. An algorithm for off-line computer-assisted seizure detection, based on these dynamical measures, is described. This algorithm does not depend upon a priori knowledge of the frequency, amplitude, or wave form characteristic of the ictal discharge. Application of this algorithm to
and clinical Neurophysiology 95 (1995) 15P-41P the detection of partial seizures in subdural and depth recordings is demonstrated. Preliminary studies indicate that this algorithm may be useful in detecting or localizing ictal discharges recorded during long term presurgical monitoring procedures. Although computationally intense, the algorithm can be employed off-line using a personal computer with a 486 microprocessor. Supported
by NIH Grant ROl NS31451-02.
19. Effects of nitrous oxide (N,O) on interictal spikes during electrocorticography. - S. Hosain, L. Nagarqjan, R. Fraser, A. Van Poznak and D. Labar (New York Hospital, Cornell Medical Center, New York, NY) Interictal spikes on intraoperative electrocorticography (ECOG) during epilepsy surgery are used to tailor resections. There have been few systematic quantitative studies of anesthetic effects on these interictal discharges, and specifically none on the effects of N,O. We calculated the spike rate on intraoperative ECoG from 2 min sample with the patient on N,O (expired concentration > 70%) and then 2 min sample with the patient off N,O (expired concentration < 10%). We studied 17 epilepsy surgery patients with subdural strips or grids analyzing the most active recording contact. The inspired concentration of isoflurane was maintained between 0.25 and 0.5%. Nine patients had fewer spikes on N,O, while 8 patients had more spikes on N,O. The mean spike rate on N,O was 42.44/min and 41_56/min when N,O was off. These rates were not significantly different (Wilcoxon matched pairs sign rank test). Our study indicates that N,O at doses employed does not effect interictal ECoG spikes, and this agent may be used during epilepsy surgery without concern about suppression of epileptiform activity in seizure 20. Predictive value of intraoperative electrocorticography outcome after temporal lobectomy. - L. De Paola, J.R. Gates, P.E. Penovich, F.J. Ritter, M.D. Frost, M.E. Dunn and W. Courtney (The Minnesota Epilepsy Group of United and Children’s Hospital, St. Paul, MN) Objective: corticography
to determine the predictive value of intraoperative electro(IECOG) in seizure outcome after temporal lobectomy
(TL). Methods: inclusion criteria: (a) patients subjected to TL with a minimum of 6 months of follow-up; (b> absence of space occupying lesions demonstrated by MRI; (c) IECoG performed during a minimum of 5-15 min after the point in which maximal resection was achieved, as recommended by IECoG or limited by functional anatomy. Expired isoforane was kept below 0.1%. Recordings were obtained from the mesial and lateral aspects of the temporal lobe. The presence (“positive”) or absence (“negative”) of epileptiform activity (EA) and its relation with seizure outcome was analyzed with a chi-square analysis. Results: 55 patients (66 surgeries) met the above criteria. The overall results demonstrated worthwhile results in 70% of the patients (Engel’s classes I (62%) and II (8%)). IECoG was performed in 59 surgeries, being “positive” in 28 (16 classes I, II/12 classes III, IV) and “negative” in 31 (21 class I and 10 classes III, IV). A statistically significant relation between EA during IECoG and seizure outcome after TL was not demonstrated (P = 0.40). Conclusion: our results suggest a limited value for IECoG as predictor of seizure outcome after TL, beyond the point of IECoG directed maximal resection. 21. Intracranial EEG validation of spike propagation predicted by dipole models. - J.S. Ebersole, S. Hawes and M. Scherg ’ (VA Medical Center, West Haven, CT, and a University of Heidelberg, Heidelberg, Germany) Variable recruitment of cortex surrounding the spike onset zone produces different generator configurations, which may falsely give the
Society Proceedings/Electroencephalography
16. EEG spectral findings in healthy, elderly men and women - sex differences. - R.P. Brenner qb, RF. Ulrich ’ and C.F. Reynolds, III a,’ (Departments of ’ Psychiatry and b Neurology, University of Pittsburgh, School of Medicine, Pittsburgh, PA) There are few quantitative EEG studies evaluating gender-related differences in healthy, elderly subjects. Therefore, we performed computerized spectral analysis of the EEG in 119 healthy, elderly men and women, ranging in age from 60 to 87 years with a mean age of 70.4 years, to evaluate gender-related variance. The groups did not differ in age, education levels or Folstein score. We compared men (n = 53) to women (n = 661, controlling for the effect of age by decade. Compared to men, women had a significant (usually P < 0.01) increase in the parasagittal mean frequency (l-30 Hz and 4-20 Hz), beta, and beta,, while alpha, and theta-beta were decreased. Although the mean frequency was lower in older subjects, there was no significant difference by decade. In addition, there were no significant sex-age interactions. Our study adds to the existing literature of adult spectral EEG gender-related differences and extends it to late-life subjects. Although the basis for these differences is uncertain, we feel that quantitative EEG must control for variance related to gender. 17. Spatiotemporal dynamic analysis of electroconvulsive seizures. - A.D. Krystal, H.S. Greenside ‘, R.D. Weiner Gassert (Department of Psychiatry, Duke University Center, and “Departments of Physics and Computer Duke University, Durham, NC)
therapy and D. Medical Science,
The physiology of electroconvulsive therapy (ECT) seizures is poorly understood. It has been assumed that these seizures are a state of decreased complexity in brain function (neuronal hypersynchrony) compared with waking. Yet, the spatiotemporal properties of the EEG recorded during ECT seizures (ictal EEG) have not been studied quantitatively, despite evidence that ictal EEG properties relate to ECT side effects and efficacy. We preliminarily studied the spatiotemporal dynamics of 21-channel ictal EEG data recorded from 9 depressed subjects clinically referred for ECT. Coherence was calculated for multiple pairs of electrodes to reflect the pattern of the relationship of brain electrical activity across the scalp. The local largest Lyapunov exponent (LI was calculated for each channel to reveal the spatial pattern of temporal complexity across the scalp. These indices were calculated for the waking EEG recorded before the ECT course, the post-anesthesia (but pre-stimulus) EEG, and the ictal EEG. Compared with the waking and post-anesthesia states, we found a marked increase in the coherence of EEG activity across the scalp and evidence for a marked decrease in both temporal and spatial complexity of the ictal EEG (decrease in L across the whole scalp) during the seizures. 18. Computer-assisted seizure detection based on quantitative dynamical measures. - J.C. Sackellares, L.D. Iasemidis, A. Berreto, R.L. Gilmore, R.A. Savit ‘, S.A. Roper and B.M. Uthman (Departments of Neurology, Neurosurgery and Electrical Engineering, University of Florida, Gainesville, FL, and * Department of Physics, University of Michigan, Ann Arbor, MI) Dynamical measures developed to study complex systems provide a quantitative description of the time-varying properties of the EEG. We have found characteristic changes in dynamical measures such as the Lyapunov exponent and time dependency indices during partial seizures of temporal lobe origin. The Lyapunov exponent reflects the chaoticity of the signal. The time dependency indices reflect the temporal structure of the signal. An algorithm for off-line computer-assisted seizure detection, based on these dynamical measures, is described. This algorithm does not depend upon a priori knowledge of the frequency, amplitude, or wave form characteristic of the ictal discharge. Application of this algorithm to
and clinical Neurophysiology 95 (1995) 15P-41P the detection of partial seizures in subdural and depth recordings is demonstrated. Preliminary studies indicate that this algorithm may be useful in detecting or localizing ictal discharges recorded during long term presurgical monitoring procedures. Although computationally intense, the algorithm can be employed off-line using a personal computer with a 486 microprocessor. Supported
by NIH Grant ROl NS31451-02.
19. Effects of nitrous oxide (N,O) on interictal spikes during electrocorticography. - S. Hosain, L. Nagarqjan, R. Fraser, A. Van Poznak and D. Labar (New York Hospital, Cornell Medical Center, New York, NY) Interictal spikes on intraoperative electrocorticography (ECOG) during epilepsy surgery are used to tailor resections. There have been few systematic quantitative studies of anesthetic effects on these interictal discharges, and specifically none on the effects of N,O. We calculated the spike rate on intraoperative ECoG from 2 min sample with the patient on N,O (expired concentration > 70%) and then 2 min sample with the patient off N,O (expired concentration < 10%). We studied 17 epilepsy surgery patients with subdural strips or grids analyzing the most active recording contact. The inspired concentration of isoflurane was maintained between 0.25 and 0.5%. Nine patients had fewer spikes on N,O, while 8 patients had more spikes on N,O. The mean spike rate on N,O was 42.44/min and 41_56/min when N,O was off. These rates were not significantly different (Wilcoxon matched pairs sign rank test). Our study indicates that N,O at doses employed does not effect interictal ECoG spikes, and this agent may be used during epilepsy surgery without concern about suppression of epileptiform activity in seizure 20. Predictive value of intraoperative electrocorticography outcome after temporal lobectomy. - L. De Paola, J.R. Gates, P.E. Penovich, F.J. Ritter, M.D. Frost, M.E. Dunn and W. Courtney (The Minnesota Epilepsy Group of United and Children’s Hospital, St. Paul, MN) Objective: corticography
to determine the predictive value of intraoperative electro(IECOG) in seizure outcome after temporal lobectomy
(TL). Methods: inclusion criteria: (a) patients subjected to TL with a minimum of 6 months of follow-up; (b> absence of space occupying lesions demonstrated by MRI; (c) IECoG performed during a minimum of 5-15 min after the point in which maximal resection was achieved, as recommended by IECoG or limited by functional anatomy. Expired isoforane was kept below 0.1%. Recordings were obtained from the mesial and lateral aspects of the temporal lobe. The presence (“positive”) or absence (“negative”) of epileptiform activity (EA) and its relation with seizure outcome was analyzed with a chi-square analysis. Results: 55 patients (66 surgeries) met the above criteria. The overall results demonstrated worthwhile results in 70% of the patients (Engel’s classes I (62%) and II (8%)). IECoG was performed in 59 surgeries, being “positive” in 28 (16 classes I, II/12 classes III, IV) and “negative” in 31 (21 class I and 10 classes III, IV). A statistically significant relation between EA during IECoG and seizure outcome after TL was not demonstrated (P = 0.40). Conclusion: our results suggest a limited value for IECoG as predictor of seizure outcome after TL, beyond the point of IECoG directed maximal resection. 21. Intracranial EEG validation of spike propagation predicted by dipole models. - J.S. Ebersole, S. Hawes and M. Scherg ’ (VA Medical Center, West Haven, CT, and a University of Heidelberg, Heidelberg, Germany) Variable recruitment of cortex surrounding the spike onset zone produces different generator configurations, which may falsely give the