EEG-SSM: Leveraging State-Space Model for Dementia Detection

TL;DR

This work introduces EEG-SSM, a dual-temporal and spectral state-space framework for dementia classification using resting-state EEG. Built on a Mamba encoder, EEG-SSM combines a temporal stream with a spectral (PSD-based) stream and an Opt-W spectral weighting mechanism to adaptively fuse band-specific information across subjects. Across a 3-class task (HC, AD, FTD) on a 19-channel, 500 Hz EEG dataset, EEG-SSM achieves a peak accuracy of 91.0% with 2-second segments and demonstrates robustness across sampling rates and segment lengths, outperforming RNN, EEG-Net, and EEG-Transformer baselines. The approach highlights the value of integrating temporal dynamics and spectral content within a scalable, low-parameter architecture for clinical dementia screening, while noting opportunities to incorporate spatial features in future work. Overall, EEG-SSM offers a promising, efficient tool for early dementia screening with strong potential for clinical translation.

Abstract

State-space models (SSMs) have garnered attention for effectively processing long data sequences, reducing the need to segment time series into shorter intervals for model training and inference. Traditionally, SSMs capture only the temporal dynamics of time series data, omitting the equally critical spectral features. This study introduces EEG-SSM, a novel state-space model-based approach for dementia classification using EEG data. Our model features two primary innovations: EEG-SSM temporal and EEG-SSM spectral components. The temporal component is designed to efficiently process EEG sequences of varying lengths, while the spectral component enhances the model by integrating frequency-domain information from EEG signals. The synergy of these components allows EEG-SSM to adeptly manage the complexities of multivariate EEG data, significantly improving accuracy and stability across different temporal resolutions. Demonstrating a remarkable 91.0 percent accuracy in classifying Healthy Control (HC), Frontotemporal Dementia (FTD), and Alzheimer's Disease (AD) groups, EEG-SSM outperforms existing models on the same dataset. The development of EEG-SSM represents an improvement in the use of state-space models for screening dementia, offering more precise and cost-effective tools for clinical neuroscience.

Figures (3)

• Figure 1: Overview of our proposed framework.
• Figure 2: Bottleneck structure for Adaptive EEG Wavelet-Specific Weight Module with three classes and R wavelet-specific models.
• Figure 3: Confusion matrix displaying classification results on the test set using 2-second EEG data segments.