Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ConvMambaNet: A Hybrid CNN-Mamba State Space Architecture for Accurate and Real-Time EEG Seizure Detection

Md. Nishan Khan, Kazi Shahriar Sanjid, Md. Tanzim Hossain, Asib Mostakim Fony, Istiak Ahmed, M. Monir Uddin

TL;DR

ConvMambaNet addresses the need for accurate and real-time EEG seizure detection by fusing CNN-based spatial feature extraction with Mamba-SSM for long-range temporal modeling. The hybrid architecture delivers state-of-the-art performance on CHB-MIT, achieving an accuracy of around $99\%$, with near-perfect precision and recall for non-seizure events and an AUC of approximately $0.97$. Key contributions include end-to-end trainability, stability-focused initializations, and attention-enhanced temporal processing that outperform CNN, RNN, and Transformer baselines while maintaining real-time efficiency. This work has practical impact for clinical EEG monitoring by offering a scalable, low-latency solution capable of handling heavy class imbalance and cross-subject variability.

Abstract

Epilepsy is a chronic neurological disorder marked by recurrent seizures that can severely impact quality of life. Electroencephalography (EEG) remains the primary tool for monitoring neural activity and detecting seizures, yet automated analysis remains challenging due to the temporal complexity of EEG signals. This study introduces ConvMambaNet, a hybrid deep learning model that integrates Convolutional Neural Networks (CNNs) with the Mamba Structured State Space Model (SSM) to enhance temporal feature extraction. By embedding the Mamba-SSM block within a CNN framework, the model effectively captures both spatial and long-range temporal dynamics. Evaluated on the CHB-MIT Scalp EEG dataset, ConvMambaNet achieved a 99% accuracy and demonstrated robust performance under severe class imbalance. These results underscore the model's potential for precise and efficient seizure detection, offering a viable path toward real-time, automated epilepsy monitoring in clinical environments.

ConvMambaNet: A Hybrid CNN-Mamba State Space Architecture for Accurate and Real-Time EEG Seizure Detection

TL;DR

ConvMambaNet addresses the need for accurate and real-time EEG seizure detection by fusing CNN-based spatial feature extraction with Mamba-SSM for long-range temporal modeling. The hybrid architecture delivers state-of-the-art performance on CHB-MIT, achieving an accuracy of around , with near-perfect precision and recall for non-seizure events and an AUC of approximately . Key contributions include end-to-end trainability, stability-focused initializations, and attention-enhanced temporal processing that outperform CNN, RNN, and Transformer baselines while maintaining real-time efficiency. This work has practical impact for clinical EEG monitoring by offering a scalable, low-latency solution capable of handling heavy class imbalance and cross-subject variability.

Abstract

Epilepsy is a chronic neurological disorder marked by recurrent seizures that can severely impact quality of life. Electroencephalography (EEG) remains the primary tool for monitoring neural activity and detecting seizures, yet automated analysis remains challenging due to the temporal complexity of EEG signals. This study introduces ConvMambaNet, a hybrid deep learning model that integrates Convolutional Neural Networks (CNNs) with the Mamba Structured State Space Model (SSM) to enhance temporal feature extraction. By embedding the Mamba-SSM block within a CNN framework, the model effectively captures both spatial and long-range temporal dynamics. Evaluated on the CHB-MIT Scalp EEG dataset, ConvMambaNet achieved a 99% accuracy and demonstrated robust performance under severe class imbalance. These results underscore the model's potential for precise and efficient seizure detection, offering a viable path toward real-time, automated epilepsy monitoring in clinical environments.
Paper Structure (20 sections, 3 equations, 5 figures, 2 tables, 1 algorithm)

This paper contains 20 sections, 3 equations, 5 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Traditional Machine Learning Methods
  4. Deep Learning Advances
  5. State Space Models (SSMs)
  6. The Mamba SSM Architecture
  7. Recent CHB-MIT Seizure Detection Studies
  8. Methodology
  9. Enhanced ConvMambaNet Model Integration
  10. Hybrid Framework for EEG Analysis
  11. Experiments
  12. Dataset
  13. Data Preprocessing
  14. Evaluation Metrics
  15. Numerical Analysis
  16. ...and 5 more sections

Figures (5)

  • Figure 1: Overview of ConvMambaNet architecture combining CNN feature extraction with Mamba SSM for temporal modeling and classification. The Mamba SSM module employs selective state-space transformations with gating mechanisms to capture long-range sequential dependencies.
  • Figure 2: EEG Signals with Multiple Channels
  • Figure 3: Training and Validation Loss
  • Figure 4: Training and Validation Accuracy
  • Figure 5: (a) Confusion matrix showing class-level prediction accuracy, and (b) ROC curve illustrating the model's discriminative power.