Abstracts

Dominant Harmonic Pattern as an Ictal Marker of the Epileptogenic Zone in Focal Epilepsy: A Stereo-EEG Study

Abstract number : 1.187
Submission category : 3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year : 2023
Submission ID : 361
Source : www.aesnet.org
Presentation date : 12/2/2023 12:00:00 AM
Published date :

Authors :
Presenting Author: Lingqi Ye, MD – Department of Neurology, Epilepsy Center, Second Affiliated Hospital Zhejiang University School of Medicine

Lingli Hu, None – Department of Neurology, Epilepsy Center, Second Affiliated Hospital, School of Medicine, Zhejiang University; Hongyi Ye, None – Department of Neurology, Epilepsy Center, Second Affiliated Hospital, School of Medicine, Zhejiang University; Cong Chen, None – Department of Neurology, Epilepsy Center, Second Affiliated Hospital, School of Medicine, Zhejiang University; Shuang Wang, None – Department of Neurology, Epilepsy Center, Second Affiliated Hospital, School of Medicine, Zhejiang University

Rationale: The ictal harmonic pattern (H pattern) generated by non-linear interaction between frequencies may embrace a reservoir of localization value for the epileptogenic zone (EZ). This study aims to determine whether dominant ictal H pattern and its features can optimally define the EZ and predict surgical outcome.

Methods: A cohort of 134 patients who had undergo stereo-electroencephalography (SEEG) evaluation were reviewed and 114 of them exhibited ictal H pattern on time-frequency plots. After screened for inclusion and exclusion criteria, we extracted features of H pattern using time-frequency analysis of ictal SEEG recordings from 85 patients with drug-resistant focal epilepsy. A threshold of Q3 was applied to discriminate Channel with dominant H pattern (ChanneldH pattern) and Channel with non-dominant H pattern (Channelnon-dH pattern). Extent of removal of the seizure-onset zone (SOZ), ChanneldH pattern and Channelnon-dH pattern was correlated with surgical outcome, classified as seizure-free (SF, Engel class Ia, n = 52) or non-seizure-free (NSF, > Engel class Ia, n = 33). The Kaplan-Meier (KM) curves were generated to visualize cumulative probability of seizure freedom. The ROC curves were built to compare the performances in predicting surgical outcome. A multivariable regression analysis identified independent predictors for seizure freedom.

Results: The ictal H pattern had a significantly higher occurrence ratio, maximal frequency, and band number in SOZ compared to early propagation zone (PZ) (p < 0.0001; p < 0.0001; p < 0.0001) and in particular, ChanneldH pattern occurred more frequently within SOZ (p < 0.0001). No relationships between the border clarity at different degree or different shapes of ictal H pattern and surgical outcome were found. The resected ratios of all ChanneldH pattern and ChanneldH pattern within non-SOZ were significantly higher in SF group (p = 0.014). Patients having received complete removal of all ChanneldH pattern and ChanneldH pattern within non-SOZ were more likely to be SF in the long-term follow-up (p = 0.008; p = 0.028). The ROC curves for seizure freedom implied the highest AUC values were 0.660 (ChannelDB within non-SOZ), which was modestly better than the predictive performance of the SOZ (0.574). Multivariable analysis identified complete removal of all ChanneldH pattern [p = 0.020, OR (95% CI) = 6.228 (1.332-29.131)], postsurgical histopathology of focal cortical dysplasia, and less implanted electrodes as independent predictors of seizure freedom.

Conclusions: Our results demonstrated the localizing value of ictal H pattern, that complete removal of brain regions expressing dominant ictal H pattern was associated with favorable surgical outcome. In addition, the time-frequency feature is easily accessible. Incorporation of H pattern into presurgical evaluation can practically aid in making surgery plans.

Funding: This work was supported by the National Natural Science Foundation of China (grant nos. 81971207, 82171437, 81971208, 82001365, and 82171889) and the Natural Science Foundation of Zhejiang Province (grant no. LD22H310003).

Neurophysiology