Graph Theory Analysis of High-Density EEG Delta-Band Lagged Coherence Shows Differences Between Focal Epilepsy and Controls
Abstract number :
2.072
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2021
Submission ID :
1826343
Source :
www.aesnet.org
Presentation date :
12/5/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:53 AM
Authors :
Brinda Sevak, MS - University of Wisconsin Madison; Klevest Gjini, MD, PhD - Researcher, Department of Neurology, University of Wisconsin Madison; Elena Monai, MD - Research Specialist, Department of Neurology, University of Wisconsin-Madison; Mariel Kalkach, MD - Research Specialist, Department of Neurology, University of Wisconsin-Madison; Melanie Boly, MD, PhD - Assistant Professor, Department of Neurology, University of Wisconsin-Madison; Aaron Struck, MD - Assistant Professor, Department of Neurology, University of Wisconsin-Madison; William S. Middleton Veterans Administration Hospital
Rationale: The brain is a network of interconnected neurons. It is hypothesized that many brain functions are a result of emergent properties of brain networks. Similarly, neurological disorders may also arise from disruption of brain networks. In fact, the network hypothesis of epilepsy speculates that aberrant network connectivity may be an essential factor leading to epilepsy. Scalp high-density EEG (hdEEG) is a noninvasive method used to quantify the macro-scale brain connectivity with a higher temporal resolution than fMRI. Characterizing the alteration of the functional connectivity of the brain may provide insight into the genesis of epilepsy and in the development of practical biomarkers. In the current study, we hypothesize that alterations in the delta band network connectivity are reflective of epilepsy. Specifically, that in focal epilepsy the global graph-theory measures of modularity and small-world index are increased as a result of the increased delta band connectivity of the “primary epileptic-network” while the global transitivity is decreased due to less organized connectivity as a secondary influence of the epileptic network.
Methods: A total of 12 patients with focal epilepsy and 11 controls were included in the study. Resting state eyes-closed scalp hdEEG data were used to estimate the cortical source signals by a "depth-weighted" linear L2-minimum norm current density estimation method followed by Destrieux atlas-based parcellation in Brainstorm. Finally, functional connectivity matrices of delta band lagged coherence were created using spectral densities obtained from Fourier transform of the parcellation-based source signals. After this, a 50% threshold was applied to the functional matrices to perform graph theory analysis. Global graph theory metrics of network modularity, transitivity and small-world index were calculated using the brain connectivity toolbox. Independent samples t-tests were performed on the obtained metrics from the epilepsy and control data to understand the differences between groups and controlled for family-wise error.
Results: Graph theory analysis on the delta band lagged coherence showed an increase in the modularity (p = 0.006) and small world index (p = 0.02) for patients with focal epilepsy as well as a decrease in the transitivity (p = 0.02) compared to the control subjects.
Conclusions: The preliminary results from the comparison of data from 12 patients and 11 controls show that there is an increase in modularity and small world index suggestive of a highly connected “epileptic network.” Additionally, there was decreased transitivity suggestive of a disruption of the global brain network function. Further study will aim at validation and association with cognitive outcomes, seizure frequency, drug-resistance and epilepsy localization.
Funding: Please list any funding that was received in support of this abstract.: JMECP R01NS1111022.
Neurophysiology