Normative Mapping of Intracranial EEG Networks for Identifying Epileptogenic Networks
Abstract number :
1.199
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2022
Submission ID :
2204118
Source :
www.aesnet.org
Presentation date :
12/3/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:23 AM
Authors :
Peter Taylor, PhD – Newcastle University; Gabrielle Schroeder, PhD – Newcastle University; Yujiang Wang, PhD – Newcastle University
Rationale: Focal epilepsy is associated not only with localised brain abnormalities, but also large-scale network alterations. In particular, abnormal network interactions may help localise the epileptogenic zone, or the patient-specific region that is indispensable for generating seizures. However, identifying reliable biomarkers for the epileptogenic zone remains an open challenge, in part because abnormal dynamics are difficult to distinguish from normal spatial heterogeneity in brain activity.
Methods: We created normative maps of intracranial EEG (iEEG) functional networks using data from 247 subjects from the Restoring Active Memory dataset. Functional networks were defined as the pairwise Pearson correlation between iEEG time series in delta, theta, alpha, beta, gamma, and broadband frequency bands. Each network was constructed using 23,248 iEEG contacts from outside the seizure onset and initial propagation zones, allowing us to approximate normal network dynamics. To aggregate functional networks across patients with different electrode implantations, iEEG contacts were assigned to regions of interest (ROIs), yielding functional network matrices with 82 ROIs for each frequency band in each subject. This approach created a distribution of expected, normative values for each functional network edge, which were summarised using the distribution's mean and standard deviation.
Results: We present normative network maps that capture normal spatial variability in iEEG functional networks. Our normative networks have properties such as high modularity, in agreement with previous studies of large-scale brain networks. Finally, we demonstrate their potential use as baselines for identifying epileptogenic networks in individual patients with focal epilepsy.
Conclusions: Our normative networks will serve as a valuable tool for uncovering pathological network interactions in patients with focal epilepsy. Future work will apply these normative networks to presurgical iEEG networks, with the goal of revealing patient-specific network abnormalities that localise the epileptogenic zone and predict treatment outcomes.
Funding: P.N.T. is supported by a UKRI Future Leaders Fellowship (MR/T04294X/1).
Neurophysiology