This is the authors' abstract. We don't add key points for this paper.
Epilepsy is a neurological disease caused by excessive neuron excitability in the brain. Feature-extracting strategies are necessary for EEG detection-dependent epilepsy to perform well. In this paper, we proposed Chi-square feature-ranking method on 22 multimodal features extracted from epileptic seizure based on time and frequency domain characteristics, entropy-based complexity-based features, wavelet-based and a few statistical features. The multimodal features were extracted based on the temporal variation spectral changes and nonlinear dynamics of brain signals. We applied robust machine learning (ML) classifiers such as k-nearest neighbor (KNN), decision tree (DT), Bayesian classifier, support vector machine (SVM) with kernels such as radial base function and linear, eXtreme boosting linear (XGBoost-L) and eXtreme boosting tree (XGBoost-tree) methods. The performance was evaluated based on top ranked 2, 4, 8, 10, 11 and all features. Using top 2–3 ranked features, few classifiers yielded highest detection performance, while the top 11 features yielded the highest performance on all the selected classifiers. Likewise, all 22 selected features provided good predictions on a few selected classifiers. Results reveal that the top 11 ranked features among 22 computed features are good predictors. The feature ranking will further help to make decisions for clinicians and doctors for its diagnosis.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Hussain et al. (2023) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: