The Sink and Source: Mapping Epileptic Activity with Directed Connectivity
Poster No:
360
Submission Type:
Late-Breaking Abstract Submission
Authors:
Hamid Karimi-Rouzbahani1, Aileen McGonigal2
Institutions:
1Queensland Brain Institute, The University of Queensland, Brisbane, Queensland, 2University of Queensland, Brisbane, QLD
First Author:
Hamid Karimi-Rouzbahani
Queensland Brain Institute, The University of Queensland
Brisbane, Queensland
Queensland Brain Institute, The University of Queensland
Brisbane, Queensland
Co-Author:
Introduction:
Epilepsy affects over 50 million people globally, with 30% unresponsive to medication (Kwan & Brodie, 2000). For focal epilepsy, presurgical evaluations, including intracranial EEG, identify seizure-generating areas. Despite advanced imaging and clinical expertise, accurate EZ localization remains challenging.
Quantification methods analysing EEG signals show promise in EZ localization, focusing on interictal (the time between seizures) or ictal (seizure) periods. Ictal patterns are prominent in the seizure onset zone (SOZ). Interictal analysis focuses on spikes, discharges, and high-frequency oscillations.
There has been a shift from analysing univariate signals to analysing networks of brain activity for localisation (Gallagher et al., 2023). Network-based localisation, aligning with epilepsy as a network disorder, outperforms univariate methods (Bernabei et al., 2022). Directed connectivity, quantifying activity flow direction, has gained attention, particularly regarding the "interictal suppression hypothesis", where the SOZ is inhibited interictally (Matarrese et al., 2023). Ictal studies indicate dominant outward flow of signals from the SOZ (Balatskaya et al., 2020), though conflicting evidence exists (An et al., 2020; Janca et al., 2021). Discrepancies may stem from varying directed connectivity methods which rely on distinct signal aspects. Objective, data-driven approaches are therefore needed.
This study objectively determines neural activity direction in and out of the SOZ using 13 directed connectivity methods and network metrics. It tests sink/source hypotheses and evaluates localisation power, potentially informing automated EZ localisation algorithms.
Quantification methods analysing EEG signals show promise in EZ localization, focusing on interictal (the time between seizures) or ictal (seizure) periods. Ictal patterns are prominent in the seizure onset zone (SOZ). Interictal analysis focuses on spikes, discharges, and high-frequency oscillations.
There has been a shift from analysing univariate signals to analysing networks of brain activity for localisation (Gallagher et al., 2023). Network-based localisation, aligning with epilepsy as a network disorder, outperforms univariate methods (Bernabei et al., 2022). Directed connectivity, quantifying activity flow direction, has gained attention, particularly regarding the "interictal suppression hypothesis", where the SOZ is inhibited interictally (Matarrese et al., 2023). Ictal studies indicate dominant outward flow of signals from the SOZ (Balatskaya et al., 2020), though conflicting evidence exists (An et al., 2020; Janca et al., 2021). Discrepancies may stem from varying directed connectivity methods which rely on distinct signal aspects. Objective, data-driven approaches are therefore needed.
This study objectively determines neural activity direction in and out of the SOZ using 13 directed connectivity methods and network metrics. It tests sink/source hypotheses and evaluates localisation power, potentially informing automated EZ localisation algorithms.
Methods:
This study used an open multi-centre intracranial EEG dataset from 55 epilepsy patients (Bernabei et al., 2022). The dataset included both subdural grid/strip (ECoG) and SEEG recordings, with patients undergoing surgical resection or laser ablation. Interictal and ictal recordings had an average of 105.6 channels per patient.
The study aimed to determine neural activity direction using 13 directed connectivity measures from the PySpi toolbox (Cliff et al., 2023), classified into four categories. Two-second epochs were analysed from interictal and ictal recordings to capture early, mid, and late dynamics. Network analysis metrics, including in/out strength, first passage time, clustering coefficient, eccentricity, and betweenness centrality, were used to characterize node behaviour.
The "sink and source SOZ" hypotheses (Fig 1) were tested by examining in/out strength differences between SOZ and non-SOZ contacts. Multivariate pattern classification with a Random Forest classifier was used to localise the SOZ, employing 10-fold cross-validation and up-sampling to address class imbalance.
Statistical analysis utilised Bayes Factor tests to compare node metrics, localisation performance, and feature contributions.
The study aimed to determine neural activity direction using 13 directed connectivity measures from the PySpi toolbox (Cliff et al., 2023), classified into four categories. Two-second epochs were analysed from interictal and ictal recordings to capture early, mid, and late dynamics. Network analysis metrics, including in/out strength, first passage time, clustering coefficient, eccentricity, and betweenness centrality, were used to characterize node behaviour.
The "sink and source SOZ" hypotheses (Fig 1) were tested by examining in/out strength differences between SOZ and non-SOZ contacts. Multivariate pattern classification with a Random Forest classifier was used to localise the SOZ, employing 10-fold cross-validation and up-sampling to address class imbalance.
Statistical analysis utilised Bayes Factor tests to compare node metrics, localisation performance, and feature contributions.
Results:
During interictal periods, five connectivity measures showed higher in-strength in the SOZ, suggesting the SOZ acts as a "sink" (Fig 2A). Conversely, in ictal periods, two measures showed higher out-strength, indicating the SOZ as a "source" (Fig 2B).
Temporal variability analysis confirmed consistent patterns across epochs. While individual connectivity measures varied.
Network metrics effectively localised the SOZ in both interictal and ictal periods (Fig 2C), with individual patient performance varying. No significant impact from patient demographics was found. Ictal localisation was stronger in early epochs, likely due to clearer SOZ separation. Eccentricity was the most informative node metric, while DI and ANM were the strongest connectivity measures (Fig 2D).
Temporal variability analysis confirmed consistent patterns across epochs. While individual connectivity measures varied.
Network metrics effectively localised the SOZ in both interictal and ictal periods (Fig 2C), with individual patient performance varying. No significant impact from patient demographics was found. Ictal localisation was stronger in early epochs, likely due to clearer SOZ separation. Eccentricity was the most informative node metric, while DI and ANM were the strongest connectivity measures (Fig 2D).
Conclusions:
We showed the variable nature of directed connectivity measures and that they can effectively localise the SOZ in both interictal and ictal periods. These results offer new insights into the flow of neural activity in epilepsy and proposes a pipeline for localising seizure activity.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s)
Neurodevelopmental/ Early Life (eg. ADHD, autism) 1
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 2
Multivariate Approaches
Keywords:
Computational Neuroscience
Electroencephaolography (EEG)
Epilepsy
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
Provide references using APA citation style.
Balatskaya, A., Roehri, N., Lagarde, S., Pizzo, F., Medina, S., Wendling, F., Bénar, C. G., & Bartolomei, F. (2020b). The “Connectivity Epileptogenicity Index ” (cEI), a method for mapping the different seizure onset patterns in StereoElectroEncephalography recorded seizures. Clinical Neurophysiology, 131(8), 1947–1955.
Bernabei, J. M., Sinha, N., Arnold, T. C., Conrad, E., Ong, I., Pattnaik, A. R., Stein, J. M., Shinohara, R. T., Lucas, T. H., Bassett, D. S., Davis, K. A., & Litt, B. (2022). Normative intracranial EEG maps epileptogenic tissues in focal epilepsy. Brain, 145(6), 1949–1961.
Cliff, O. M., Bryant, A. G., Lizier, J. T., Tsuchiya, N., & Fulcher, B. D. (2023). Unifying pairwise interactions in complex dynamics. Nature Computational Science, 3(10), 883–893.
Gallagher, R., Sinha, N., Pattnaik, A., Ojemann, W., Lucas, A., LaRocque, J., Bernabei, J., Greenblatt, A., Sweeney, E., Chen, I., Davis, K., Conrad, E., & Litt, B. (2023). Quantifying interictal intracranial EEG to predict focal epilepsy. ArXiv, arXiv:2307.15170.
Kwan, P., & Brodie, M. J. (2000). Early Identification of Refractory Epilepsy. New England Journal of Medicine, 342(5), 314–319.
Matarrese, M. A. G., Loppini, A., Fabbri, L., Tamilia, E., Perry, M. S., Madsen, J. R., Bolton, J., Stone, S. S. D., Pearl, P. L., Filippi, S., & Papadelis, C. (2023). Spike propagation mapping reveals effective connectivity and predicts surgical outcome in epilepsy. Brain, 146(9), 3898–3912.