Neural Memory Networks for Seizure Type Classification

TL;DR

This work addresses the challenge of automated seizure-type classification, leveraging the TUH EEG Seizure Corpus to benchmark cross-patient performance. It introduces a neural memory network (NMN) with an external memory module and trainable neural plasticity, combining two stacked LSTMs for short-term encoding with a memory stack $M \in \mathbb{R}^{l\times k}$ and read/write/update controllers to capture long-range dependencies; memory access is augmented with Hebbian plasticity, enabling flexible, context-dependent information retrieval. In extensive 5-fold cross-validation, the Plastic NMN achieves a weighted F1 score of $0.945$, outperforming baselines including SAE, CNNs, RNNs, RCNNs, and SeizureNet, with memory embeddings providing clearer class separation in PCA visualizations. The approach demonstrates the potential of memory-augmented architectures to handle intersubject and intrasubject variability in EEG-based seizure classification, offering a path toward more robust automated support for clinicians and epilepsy research; future work includes applying the memory component directly to raw intracranial EEG without FFT preprocessing.

Abstract

Classification of seizure type is a key step in the clinical process for evaluating an individual who presents with seizures. It determines the course of clinical diagnosis and treatment, and its impact stretches beyond the clinical domain to epilepsy research and the development of novel therapies. Automated identification of seizure type may facilitate understanding of the disease, and seizure detection and prediction has been the focus of recent research that has sought to exploit the benefits of machine learning and deep learning architectures. Nevertheless, there is not yet a definitive solution for automating the classification of seizure type, a task that must currently be performed by an expert epileptologist. Inspired by recent advances in neural memory networks (NMNs), we introduce a novel approach for the classification of seizure type using electrophysiological data. We first explore the performance of traditional deep learning techniques which use convolutional and recurrent neural networks, and enhance these architectures by using external memory modules with trainable neural plasticity. We show that our model achieves a state-of-the-art weighted F1 score of 0.945 for seizure type classification on the TUH EEG Seizure Corpus with the IBM TUSZ preprocessed data. This work highlights the potential of neural memory networks to support the field of epilepsy research, along with biomedical research and signal analysis more broadly.

Figures (4)

• Figure 1: Overview of the framework proposed for classifying seizure types using sequential and neural memory networks. 1. We use the TUH EEG Seizure Corpus which contains scalp EEG data from seizure recordings and a pre-processing strategy based on the fast Fourier transform. 2. We map each data sample with 2 stacked LSTMs as input to the memory model. 3. External memory model: The state of the memory at time instant $t-1$ is $M_{t-1}$. The input controller receives the encoded hidden states $x_t$ and determines what facts within the input data to use to query the memory $q_t$. An attention score vector $z_t$ is used to quantify the similarity between the content stored in each slot of $M_{t-1}$ and the query vector $q_t$ to generate the input to the output controller. The output controller regulates what results from the memory stack ($c_t$) are passed out to the memory module for the current state ($m_t$). The update controller updates the memory state based on the output of the memory module and propagates it to the next time step. These controllers utilise a combination of fixed weights and plastic components. 4. The output of the memory model is fed to a dense layer with a soft-max activation to predict each seizure type.
• Figure 2: Normalized confusion matrices for seizure type classification on the TUH EEG Seizure Corpus for the proposed Plastic NMN model.
• Figure 3: 2D illustration of extracted embeddings from the CNN-LSTM model for randomly selected 500 samples from the test set.
• Figure 4: 2D illustration of extracted memory embeddings from the Plastic NMN for randomly selected 500 samples from the test set.