Related Experiment Video
Updated: Jun 7, 2025
12:10
Performing Behavioral Tasks in Subjects with Intracranial Electrodes
Published on: October 2, 2014
11.3K
Research progress of epileptic seizure prediction methods based on EEG
Zhongpeng Wang1,2, Xiaoxin Song1, Long Chen1,2
1Academy of Medical Engineering and Translational Medicine, Tianjin University, Tianjin, 300072 China.
Cognitive Neurodynamics
|November 18, 2024
Summary
Predicting epileptic seizures using electroencephalography (EEG) is crucial for refractory epilepsy patients. This review explores EEG-based methods, their limitations, and the potential of deep learning for improved seizure prediction and patient quality of life.
Area of Science:
- Neurology
- Biomedical Engineering
- Artificial Intelligence
Background:
- Refractory epilepsy affects over 30% of patients globally, with unpredictable seizures impacting quality of life and safety.
- Early seizure prediction and intervention are vital for improving patient outcomes and well-being.
Purpose of the Study:
- To review and evaluate electroencephalography (EEG)-based seizure prediction methods.
- To analyze the current state of seizure prediction research using scalp and intracranial EEG.
- To identify limitations hindering clinical application and suggest future research directions, including deep learning.
Main Methods:
- Introduction to EEG-based seizure prediction system design.
- Summary of common preprocessing, feature extraction, classification, and post-processing techniques.
- Review and comprehensive evaluation of feature analysis methods for scalp and intracranial EEG.
Main Results:
- Analysis of current epileptic seizure prediction research status based on five feature analysis methods.
- Evaluation of scalp EEG versus intracranial EEG for seizure prediction.
- Identification of reasons for the lack of clinical applicability of current algorithms.
Conclusions:
- Current seizure prediction algorithms face limitations preventing clinical translation.
- Deep learning offers promising advancements over traditional machine learning for epilepsy seizure prediction.
- Further research is needed to overcome limitations and achieve breakthroughs in clinical seizure prediction.
