Identifying Discriminative Regions in Epilepsy Through Effective Connectivity Analysis
Poster No:
1252
Submission Type:
Abstract Submission
Authors:
Sina Sabzevar1, saeed masoudnia1, Mohammad-Reza Nazem-Zadeh2
Institutions:
1University of Tehran, Tehran, Tehran, 2Monash University, Melbourne, VIC
First Author:
Co-Author(s):
Introduction:
Epilepsy is a widespread chronic neurological condition that impacts people across all age groups, genders, ethnicities, social backgrounds, and regions worldwide[1]. Epilepsy disrupts specific areas of the brain, causing seizures and imposing significant neurobiological, cognitive, psychological, and social challenges on patients and their families[2]. Thus, it is crucial to identify objective, epilepsy-specific biological biomarkers. Neuroimaging offers a promising pathway for discovering such biomarkers, potentially leading to more accurate diagnoses and targeted treatments.
Methods:
Identifying brain regions that reliably distinguish epilepsy patients from healthy controls is crucial for advancing our understanding of the disorder and improving diagnostic accuracy. Leveraging resting-state functional MRI(rs-fMRI)data, this study employs a comprehensive approach that integrates causal discovery methods and advanced node embedding techniques. We processed rs-fMRI data 35 patients with unilateral TLE (21 with left TLE and 14 with right TLE), as well as 11 healthy control subjects. MRI data were collected using a 3-T Siemens Prisma scanner at the Iranian National Brain Mapping Laboratory (NMBL), focusing on 116 brain regions based on the AAL116 atlas. The causal discovery method, RDCM[3], was used to uncover directional relationships (effective connectivity) between areas, while Node2vec[4] generated feature vectors that captured complex network properties. To address the class imbalance, we employed an up-sampling technique, SMOTE[5], followed by Support Vector Machine (SVM) classification with Leave-One-out Cross Validation for every region to find the most discriminative regions, offering potential biomarkers of epilepsy.
Results:
After identifying the Regions of Interest (ROIs) and extracting time series data from each region, we performed an Effective Connectivity analysis using Regression Dynamic Causal Modeling to evaluate the interactions among the ROIs. We then applied Node2Vec to convert the graph data into feature vectors. To ensure a balanced dataset, we utilized the Synthetic Minority Oversampling Technique (SMOTE) on training data, which improves model generalization and prevents the classifier from favoring a specific class. Finally, by applying Support Vector Machine (SVM) with Leave-One-Out Cross Validation, we achieved 94.68% sensitivity, 82.50% specificity, 91.84% accuracy, and a 0.87 ROC-AUC. Figure 1 highlights the most distinctive regions associated with epilepsy, while Figure 2 demonstrates the effective connectivity strength in epilepsy compared to healthy control subjects.
·Figure 1: This plot visualizes the brain regions most affected by temporal lobe epilepsy, based on recent literature. The highlighted regions include the hippocampus (HIP), amygdala (AMYG), olfactory
·Figure 2: This circular graph illustrates the differences in effective connectivity patterns between healthy and epilepsy subjects. The visualization highlights disrupted or altered connections in key
Conclusions:
In this study, we identified key brain regions that play a discriminative role in differentiating epilepsy patients from healthy controls. Our analysis revealed significant differences in effective connectivity patterns between these groups, as illustrated in our comparative plots in FIG.2. By combining advanced feature extraction technique, Node2Vec with dataset balancing via SMOTE, and employing an SVM classifier with Leave-One-Out Cross Validation, we achieved high sensitivity, specificity, accuracy, and ROC-AUC scores. These findings highlight the potential of connectivity-based approaches in improving the diagnosis and understanding of epilepsy.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s)
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Classification and Predictive Modeling 2
Connectivity (eg. functional, effective, structural) 1
Neuroinformatics and Data Sharing:
Brain Atlases
Keywords:
Epilepsy
FUNCTIONAL MRI
Machine Learning
Other - effective connectivity
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:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
[2] A. Fallahi, M. Pooyan, J. M. Habibabadi, and M.-R. Nazem-Zadeh, "Comparison of multimodal findings on epileptogenic side in temporal lobe epilepsy using self-organizing maps," Magnetic Resonance Materials in Physics, Biology and Medicine, pp. 1-18, 2021.
[3] S. Frässle et al., "Regression dynamic causal modeling for resting-state fMRI," (in eng), Hum Brain Mapp, vol. 42, no. 7, pp. 2159-2180, May 2021, doi: 10.1002/hbm.25357.
[4] A. Grover and J. Leskovec, "node2vec: Scalable Feature Learning for Networks," (in eng), Kdd, vol. 2016, pp. 855-864, Aug 2016, doi: 10.1145/2939672.2939754.
[5] N. V. Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer, "SMOTE: synthetic minority over-sampling technique," J. Artif. Int. Res., vol. 16, no. 1, pp. 321–357, 2002.