Brain Network in Epilepsy » Department of Neurology » College of Medicine

Brain Network

Our research explores the intricate brain network dynamics associated with epilepsy and leverages these insights to advance diagnostic and therapeutic strategies

We introduced a “genetic fingerprinting of brain networks” by identifying heritable traits within resting-state network topology derived from MEG and fMRI (Pourmotabbed et al., 2024). This work reveals genetic influences on connectivity patterns, paving the way for personalized medicine in neurological conditions.

Heritability of Local Brain Networks. Heritability scores (h²) of local brain network measures (strength, eigenvector centrality, clustering coefficient, and nodal efficiency) in resting-state MEG (Top) and fMRI (Bottom). MEG consistently demonstrated higher heritability scores than fMRI, highlighting its potential in genetic studies of brain connectivity. (Pourmotabbed et al., 2024)

**Performance of Genetic Fingerprinting Algorithm.** Violin plots show the algorithm’s area under the receiver operating curve (AUC) and accuracy in differentiating monozygotic twins. Genetic fingerprinting using MEG outperformed fMRI with global graph measures, while comparable accuracy was achieved with local graph properties. (Pourmotabbed et al., 2024)
**dwPLI**: Debiased Weighted Phase Lag Index; **AEC**: Amplitude Envelope Correlation; **Cor**: Correlation

We assessed the reproducibility of graph-based metrics in MEG-derived functional connectivity, examining both sensor and source spaces. Our findings confirm the stability of these metrics, reinforcing the reliability of MEG connectivity data for future clinical and research applications (Pourmotabbed et al., 2022).

**Reproducibility of MEG Resting-State Connectivity Metrics**. Test-retest reliability of MEG graph measures in resting-state data, evaluated with the intraclass correlation coefficient (ICC) for two connectivity measures (dwPLI and AEC) across frequency bands. Notably, MEG metrics showed good to excellent reliability (0.6 ≤ ICC < 0.75 and ICC ≥ 0.75). (Pourmotabbed et al., 2022)
**dwPLI**: debiased weighted phase lag index; **AEC**: amplitude envelope correlation; **lBeta**: low beta band (13-20 Hz); **hBeta**: high beta band (20-30 Hz); **lGamma**: low gamma band (30-50 Hz).

Our epilepsy research focuses on how complex brain region interactions contribute to epileptic activity. Using MEG and intracranial EEG, we map connectivity patterns that help pinpoint epilepsy’s origins and lateralization. Through resting-state MEG analysis, we uncover inter- and intra-hemispheric connectivity differences, providing essential insights for lateralizing epilepsy in individual patients (Pourmotabbed et al., 2020).

**Distinctive Network Differences in Focal Epilepsy.** Sub-network showing significant differences between patients with right-hemispheric and left-hemispheric focal epilepsy in the theta band. (Pourmotabbed et al., 2020)

We also developed predictive models for seizure outcomes and treatment responses in patients undergoing vagus nerve stimulation (VNS) by analyzing network topology in preimplantation MEG data. This predictive approach aids in tailoring VNS treatment strategies for epilepsy patients (Babajani-Feremi et al., 2018).

**Brain Network Modularity in VNS Responders and Non-Responders.** Modularity of the MEG resting-state brain network in three frequency bands for healthy controls, VNS responders, and non-responders, illustrating network organization differences linked to treatment outcomes. (Babajani-Feremi et al., 2018)

Related Publications

Genetic fingerprinting with heritable phenotypes of the resting-state brain network topology

H. Pourmotabbed, D. F. Clarke, C. Chang, and A. Babajani-Feremi, “Genetic fingerprinting with heritable phenotypes of the resting-state brain network topology,” Commun Biol, vol. 7, no. 1, p. 1221, Sep 30, 2024, doi: https://doi.org/10.1038/s42003-024-06807-0.

An application of dynamical directed connectivity of ictal intracranial EEG recordings in seizure onset zone localization

M. Nahvi, G. Ardeshir, M. Ezoji, A. Tafakhori, S. Shafiee, and A. Babajani-Feremi, “An application of dynamical directed connectivity of ictal intracranial EEG recordings in seizure onset zone localization,” J Neurosci Methods, vol. 386, p. 109775, Feb 15, 2023, doi: https://doi.org/10.1016/j.jneumeth.2022.109775.

Reproducibility of graph measures derived from resting-state MEG functional connectivity metrics in sensor and source spaces

H. Pourmotabbed, A. L. de Jongh Curry, D. F. Clarke, E. C. Tyler-Kabara, and A. Babajani-Feremi, “Reproducibility of graph measures derived from resting-state MEG functional connectivity metrics in sensor and source spaces,” Hum Brain Mapp, Jan 12, 2022, doi: https://doi.org/10.1002/hbm.25726.

Lateralization of epilepsy using intra-hemispheric brain networks based on resting-state MEG data

H. Pourmotabbed, J. W. Wheless, and A. Babajani-Feremi, “Lateralization of epilepsy using intra-hemispheric brain networks based on resting-state MEG data,” Hum Brain Mapp, vol. 41, no. 11, pp. 2964-2979, Aug 1, 2020, doi: https://doi.org/10.1002/hbm.24990.

Predicting seizure outcome of vagus nerve stimulation using MEG-based network topology

A. Babajani-Feremi, N. Noorizadeh, B. Mudigoudar, and J. W. Wheless, “Predicting seizure outcome of vagus nerve stimulation using MEG-based network topology,” Neuroimage Clin, vol. 19, pp. 990-999, 2018, doi: https://doi.org/10.1016/j.nicl.2018.06.017

Identifying seizure onset zone from electrocorticographic recordings: A machine learning approach based on phase locking value

B. Elahian, M. Yeasin, B. Mudigoudar, J. W. Wheless, and A. Babajani-Feremi, “Identifying seizure onset zone from electrocorticographic recordings: A machine learning approach based on phase locking value,” Seizure, vol. 51, pp. 35-42, Oct 2017, doi: https://doi.org/10.1016/j.seizure.2017.07.010.