Enhanced epilepsy detection using discrete wavelet transform and bandpass filtering on EEG data: integration of ART-based and LVQ models

Accurate detection of epileptic seizures from EEG signals is vital for early diagnosis and treatment of epilepsy. However, EEG signals are inherently nonstationary and noisy, posing significant challenges for classification. This study presents a lightweight and interpretable framework for epileptic seizure detection, combining Discrete Wavelet Transform (DWT) and bandpass filtering for robust time–frequency feature extraction. We evaluate multiple adaptive classifiers, including Adaptive Resonance Theory (ART1, ARTMAP) and Learning Vector Quantization (LVQ), enhanced through grid search and ensemble learning. Experiments were conducted using the Bonn EEG dataset, focusing on classifying interictal and ictal EEG signals. Among the evaluated models, ARTMAP with preprocessing and ensemble techniques achieved the best performance (AUC = 0.85), followed by ART1 with grid search and ensemble (AUC = 0.81). These results demonstrate that interpretable, adaptive models can offer competitive performance in EEG classification, while maintaining computational efficiency compared to deep learning methods. The proposed method provides a practical and transparent solution for EEG-based seizure detection, especially suited for real-time clinical applications and resource-constrained environments.