Epileptic Seizure Detection Using Machine Learning: Taxonomy, Opportunities, and Challenges.

Area of Science:

• Neurology
• Biomedical Engineering
• Data Science

Background:

• Epilepsy is a serious neurological disorder characterized by recurrent seizures.
• Automatic diagnosis of epileptic seizures is crucial for clinical management.
• Machine learning (ML) offers promising avenues for early seizure prediction.

Purpose of the Study:

• To systematically review feature selection and classification methods for epileptic seizure prediction.
• To identify the most frequently used feature extraction techniques and classifiers in EEG-based seizure detection.
• To provide a comprehensive overview of the current state-of-the-art in ML for epilepsy.

Main Methods:

• Systematic literature review of studies from major scientific repositories (MDPI, IEEE Xplore, Wiley, Elsevier, ACM, Springer Link).
• Analysis of feature extraction methods and machine learning classifiers applied to electroencephalogram (EEG) data.
• Creation of a taxonomy summarizing existing solutions and examination of benchmark datasets.

Main Results:

• Identification of prevalent feature extraction techniques and ML classifiers for epileptic seizure classification.
• Analysis of the performance of various classifiers on different datasets.
• Synthesis of current approaches in a structured taxonomy.

Conclusions:

• The review highlights key methods and classifiers for EEG-based epilepsy prediction.
• Identified research gaps, challenges, and future opportunities for enhancing seizure prediction accuracy.
• Provides a foundation for further research and development in automated epilepsy diagnosis.