Related Experiment Video
Updated: Aug 13, 2025
09:57
Author Spotlight: Advancing Pediatric Epilepsy Surgery in Children Through Novel Biomarkers and Enhanced Localization
Published on: September 20, 2024
2.7K
Can Presurgical Interhemispheric EEG Connectivity Predict Outcome in Hemispheric Surgery? A Brain Machine Learning
Chiara Pepi1, Mattia Mercier1, Giusy Carfì Pavia1
1Rare and Complex Epilepsies Unit, Department of Neuroscience, Bambino Gesù Children's Hospital, IRCCS, Full Member of European Reference Network EpiCARE, 00165 Rome, Italy.
Brain Sciences
|January 21, 2023
Summary
Machine learning accurately predicts seizure outcomes after hemispherotomy surgery. Pre-surgical EEG connectivity analysis shows potential for forecasting results in epilepsy patients.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Epileptology
Background:
- Hemispherotomy (HT) is a key surgical intervention for drug-resistant epilepsy stemming from hemispheric lesions.
- Predicting seizure outcomes post-HT remains challenging, necessitating advanced analytical methods.
Purpose of the Study:
- To investigate the predictive power of Inter-hemispheric EEG Connectivity (IC) for seizure outcomes following hemispherotomy.
- To develop and validate a machine learning (ML) model for predicting post-surgical epilepsy control.
Main Methods:
- Utilized pre-surgical EEG data from 21 pediatric patients who underwent HT.
- Employed an Artificial Neuronal Network (ANN) trained on EEG features to predict epilepsy outcomes.
- Analyzed 5-second windows of wakefulness and sleep EEG data.
Main Results:
- 71% of patients achieved seizure and drug-free status at follow-up.
- The ANN model demonstrated 73.3% accuracy in predicting outcomes.
- The model correctly classified 85% of seizure-free and 40% of non-seizure-free patients.
Conclusions:
- Pre-surgical EEG features, particularly IC, show significant potential in predicting epilepsy outcomes after HT.
- The developed ML model offers a high-accuracy computational approach to guide surgical decision-making and patient management.
