Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Patient-Independent Neonatal Seizure Prediction Model Using Reduced Montage EEG and ECG

Sithmini Ranasingha, Agasthi Haputhanthri, Hansa Marasinghe, Nima Wickramasinghe, Kithmin Wickremasinghe, Jithangi Wanigasinghe, Chamira U. S. Edussooriya, Joshua P. Kulasingham

TL;DR

This work tackles neonatal seizure prediction with a patient-independent approach by fusing reduced-montage EEG and ECG through MFCC features and a lightweight CNN with channel attention. The model achieves high predictive performance in 10-fold cross-validation on neonatal Helsinki data (and comparable results on Siena adults), with lead times up to 30 minutes before onset, and demonstrates transfer-learning–driven generalization to unseen subjects via LOPO. SHAP-based explanations provide interpretable channel-level localization of predicted preictal activity, supporting clinical trust and potential seizure focus localization. While LOPO performance improves substantially with limited fine-tuning data, the approach remains challenged by inter-subject variability, underscoring the value of transfer learning and transparent AI in real NICU deployment.

Abstract

Neonates are highly susceptible to seizures, often leading to short or long-term neurological impairments. However, clinical manifestations of neonatal seizures are subtle and often lead to misdiagnoses. This increases the risk of prolonged, untreated seizure activity and subsequent brain injury. Continuous video electroencephalogram (cEEG) monitoring is the gold standard for seizure detection. However, this is an expensive evaluation that requires expertise and time. In this study, we propose a convolutional neural network-based model for early prediction of neonatal seizures by distinguishing between interictal and preictal states of the EEG. Our model is patient-independent, enabling generalization across multiple subjects, and utilizes mel-frequency cepstral coefficient matrices extracted from multichannel EEG and electrocardiogram (ECG) signals as input features. Trained and validated on the Helsinki neonatal EEG dataset with 10-fold cross-validation, the proposed model achieved an average accuracy of 97.52%, sensitivity of 98.31%, specificity of 96.39%, and F1-score of 97.95%, enabling accurate seizure prediction up to 30 minutes before onset. The inclusion of ECG alongside EEG improved the F1-score by 1.42%, while the incorporation of an attention mechanism yielded an additional 0.5% improvement. To enhance transparency, we incorporated SHapley Additive exPlanations (SHAP) as an explainable artificial intelligence method to interpret the model and provided localization of seizure focus using scalp plots. The overall results demonstrate the model's potential for minimally supervised deployment in neonatal intensive care units, enabling timely and reliable prediction of neonatal seizures, while demonstrating strong generalization capability across unseen subjects through transfer learning.

A Patient-Independent Neonatal Seizure Prediction Model Using Reduced Montage EEG and ECG

TL;DR

This work tackles neonatal seizure prediction with a patient-independent approach by fusing reduced-montage EEG and ECG through MFCC features and a lightweight CNN with channel attention. The model achieves high predictive performance in 10-fold cross-validation on neonatal Helsinki data (and comparable results on Siena adults), with lead times up to 30 minutes before onset, and demonstrates transfer-learning–driven generalization to unseen subjects via LOPO. SHAP-based explanations provide interpretable channel-level localization of predicted preictal activity, supporting clinical trust and potential seizure focus localization. While LOPO performance improves substantially with limited fine-tuning data, the approach remains challenged by inter-subject variability, underscoring the value of transfer learning and transparent AI in real NICU deployment.

Abstract

Neonates are highly susceptible to seizures, often leading to short or long-term neurological impairments. However, clinical manifestations of neonatal seizures are subtle and often lead to misdiagnoses. This increases the risk of prolonged, untreated seizure activity and subsequent brain injury. Continuous video electroencephalogram (cEEG) monitoring is the gold standard for seizure detection. However, this is an expensive evaluation that requires expertise and time. In this study, we propose a convolutional neural network-based model for early prediction of neonatal seizures by distinguishing between interictal and preictal states of the EEG. Our model is patient-independent, enabling generalization across multiple subjects, and utilizes mel-frequency cepstral coefficient matrices extracted from multichannel EEG and electrocardiogram (ECG) signals as input features. Trained and validated on the Helsinki neonatal EEG dataset with 10-fold cross-validation, the proposed model achieved an average accuracy of 97.52%, sensitivity of 98.31%, specificity of 96.39%, and F1-score of 97.95%, enabling accurate seizure prediction up to 30 minutes before onset. The inclusion of ECG alongside EEG improved the F1-score by 1.42%, while the incorporation of an attention mechanism yielded an additional 0.5% improvement. To enhance transparency, we incorporated SHapley Additive exPlanations (SHAP) as an explainable artificial intelligence method to interpret the model and provided localization of seizure focus using scalp plots. The overall results demonstrate the model's potential for minimally supervised deployment in neonatal intensive care units, enabling timely and reliable prediction of neonatal seizures, while demonstrating strong generalization capability across unseen subjects through transfer learning.

Paper Structure

This paper contains 19 sections, 4 equations, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Data Preparation
  4. Datasets
  5. Preprocessing
  6. MFCC Calculation
  7. Proposed Model Architecture
  8. CNN Encoder
  9. Squeeze-and-Excitation Channel Attention
  10. Classifier
  11. Model Training and Interpretation
  12. 10-Fold Cross-Validation
  13. Leave-One-Patient-Out Evaluation
  14. Model Interpretation
  15. Results and Discussion
  16. ...and 4 more sections

Figures (4)

  • Figure 1: EEG and ECG recordings across four brain states associated with seizures. The figure illustrates interictal, preictal, ictal, and postictal states, along with the respective time durations considered for Helsinki dataset construction in this study.
  • Figure 2: Scalp electrode placement and EEG channel selection. Electrodes selected for this study are shaded in blue. Red and blue arrows indicate the EEG channels: red arrows correspond to the twelve channels defined in prev_fyp, while blue arrows denote the six additional channels introduced in this study.
  • Figure 3: Architecture of the proposed deep learning model for seizure prediction. Raw EEG and ECG are preprocessed, segmented into 5 s windows, and converted into MFCC matrices. Features are extracted using 2D convolutional layers with batch normalization and ReLU (blue) and max pooling layers (green). An attention mechanism precedes the dense classifier, which produces a binary output: 1 for preictal (upcoming seizure) and 0 for interictal (no seizure).
  • Figure 4: Scalp Mapping of SHAP Values for Preictal EEG Samples in Neonates. The primary seizure location is indicated below each plot. Shaded electrodes highlight the nine electrodes used in this study, while lines between electrode pairs represent the 18 EEG channels. Line intensity corresponds to SHAP-based importance, with darker lines indicating greater importance.