Composite EEG biomarker modeling and energy–accuracy trade-off analysis for multi-patient seizure detection

Deep Hybrid Neural Network for Automated Epileptic Seizure Detection from EEG Signals using MAT-LAB

A Deep Hybrid Neural Network framework that combines Convolutional Neural Networks (CNN) with the Aquila Optimizer (AO) for the automatic detection of epileptic seizures utilizing EEG data in MATLAB is introduced.

An Automated Multivariate EEG Signal Analysis Framework for Seizure Detection Using D-MVMD and BO-SVM.

The epileptic seizure (ES) is one of the most prominent neurological conditions, whose detection and classification from the electroencephalogram (EEG) signals is crucial for effective diagnosis of seizures, thereby eliminating the detrimental effects associated with it. However, the development of an automated ES detection system is hindered by the non-stationary, non-linear, and high-dimensional nature of the EEG signals, compounded by noise contamination and inter-subject variability. To address these challenges, this paper proposes an automated ES detection framework based on the De-mixing Multivariate Variational Mode Decomposition (D-MVMD) integrated with the Bayesian Optimized Support Vector Machine (BO-SVM). The D-MVMD decomposes multichannel EEG signals into band-limited intrinsic mode functions (BIMFs) while alleviating the correlation between corresponding modes through an ensemble correlation coefficient, while preserving the seizure characteristics from contamination. Multi-domain features capturing temporal, spectral, and non-linear dynamics of seizure activity are then extracted from the de-mixed BIMFs. ReliefF-ranked random forest-based feature selection is employed to find discriminative features, which are subsequently classified using the optimally tuned BO-SVM classifier. Experimental evaluation on the CHB-MIT demonstrates superior performance, with an accuracy of 98.52%, precision of 98.67%, sensitivity of 98.54%, specificity of 98.54%, and F1 score of 0.98. The model is also evaluated on the Siena dataset to assess its robustness across recording sessions of the same patient. Further, it is analyzed with other state-of-the-art methods, confirming its superior mode separation and enhanced seizure detection. Hence, this developed model proves itself to be an effective and robust model for detecting seizures using the multichannel EEG analysis.

Patient-Specific EEG Baseline Establishment Using the E-norms Method for Pediatric Seizure Detection Without Labeled Training Data

Comparison between epilepsy and migraine based on feature fusion and multiple machine learning models using electroencephalogram

Epilepsy and migraine are prone to clinical misdiagnosis due to overlapping clinical manifestations, while the limited availability of electroencephalogram (EEG) data further complicates accurate differential diagnosis. To address this challenge, this study proposes a three-class classification framework for epilepsy, migraine, and healthy controls based on EEG signals collected from 36 participants. After preprocessing the raw EEG data with a 0.5–60 Hz bandpass filter, 15 features were extracted, including 12 statistical features and three nonlinear dynamical features. Four machine learning models, namely support vector machine (SVM), random forest (RF), LightGBM, and XGBoost, were systematically evaluated for classification performance. Among them, XGBoost achieved the best overall results, with a test accuracy of 0.90 and superior performance across all major evaluation metrics. Feature importance analysis based on the gain metric of XGBoost identified the Hurst exponent as the most influential feature. Notably, its inter-group distribution differences were highly consistent with the pathological characteristics of epilepsy and migraine. These findings suggest that EEG-based feature fusion combined with machine learning provides an effective strategy for multi-class neurological disease classification, and that the Hurst exponent may serve as a promising biomarker for the differential diagnosis of epilepsy and migraine.

EEG-X: An Integrated Framework for Automated Quantitative EEG Analysis in Epilepsy Diagnosis.

Accurate interpretation of electroencephalography (EEG) remains a major challenge in epilepsy diagnosis, particularly given that patient heterogeneity hinders the cross-subject generalization of artificial intelligence based algorithms. To address the complex spatio-temporal characteristics and long-range contextual dependencies of EEG signals, we propose a novel EEG decoding frame work. The framework employs a graph convolutional neural network to capture the spatial dependencies across EEG channels, integrated with a multiresolution transformer block to capture short-term dynamics while preserving long-term contextual information. This framework is used to develop a novel seizure detection algorithm, which has been validated on a large multicenter dataset in a cross subject manner. The algorithm demonstrated superior performance across diverse patient populations, achieving an accuracy of 96.5% and a false alarm rate of 1.61/h. To facilitate downstream clinical and research applications, we establish EEG-X, a cloud-based collaborative frame work that integrates the seizure detection algorithm along with our previously established interictal epileptiform detection and EEG source imaging methods. The framework supports collaborative annotation, enables browser-based multi-platform access without requiring specific software installation, and facilitates continuous monitoring in both clinical and home settings, demonstrating the feasibility and potential efficiency gains toward the practical implementation of quantitative EEG analysis for clinicians and researchers.

Channel reduction in AdaBoost ensemble models for energy-efficient ambulatory seizure detection

Wearable electroencephalogram (EEG) devices offer promising solutions for continuous seizure monitoring in out-of-hospital settings. However, the adoption of medical-grade wearable devices remains limited by the power consumption demands of edge computing, where each active EEG channel contributes to energy expenditure through signal processing and feature extraction operations. Optimizing algorithms through feature reduction may extend battery life without compromising diagnostic reliability. This study evaluated a feature reduction strategy for a 100-tree AdaBoost ensemble model applied to binary seizure detection using the Bangalore EEG Epilepsy Dataset, comprising recordings from 60 subjects. A patient-aware data partitioning scheme (70/15/15 split) was implemented to ensure complete separation of patient identities across training, validation, and testing subsets. Feature importance scores were extracted from the trained ensemble to identify less discriminative EEG channels, and classification performance was compared between a baseline 16-channel configuration and a reduced 12-channel configuration following removal of the four least important channels. The baseline 16-channel model achieved 97.56% test accuracy, 97.29% sensitivity, 98.50% specificity, and a 2.71% seizure miss rate. Following the removal of channels x5, x9, x11, and x13, the reduced 12-channel model achieved 97.11% test accuracy, 96.71% sensitivity, 98.50% specificity, and a 3.29% seizure miss rate. The 25% reduction in active channels resulted in only a 0.45 percentage-point decrease in accuracy while yielding an estimated 33.33% theoretical extension in battery operating duration. Strategic channel reduction based on feature importance analysis can improve computational efficiency while maintaining comparable seizure detection performance. These findings support the development of energy-efficient, long-term ambulatory EEG monitoring devices that balance diagnostic reliability with extended operational duration.