NeuroSSM: Multiscale Differential State-Space Modeling for Context-Aware fMRI Analysis

TL;DR

NeuroSSM introduces a multiscale state-space model for fMRI analysis that captures both fast transient and slow global dynamics efficiently, improving sensitivity and performance over existing methods.

Contribution

The paper presents NeuroSSM, a novel multiscale differential state-space architecture for end-to-end raw fMRI signal analysis, addressing limitations of prior single-scale and connectivity-based models.

Findings

NeuroSSM achieves competitive accuracy on clinical and non-clinical datasets.

The model effectively captures multiscale temporal dynamics in fMRI signals.

NeuroSSM demonstrates improved efficiency compared to transformer-based approaches.

Abstract

Accurate fMRI analysis requires sensitivity to temporal structure across multiple scales, as BOLD signals encode cognitive processes that emerge from fast transient dynamics to slower, large-scale fluctuations. Existing deep learning (DL) approaches to temporal modeling face challenges in jointly capturing these dynamics over long fMRI time series. Among current DL models, transformers address long-range dependencies by explicitly modeling pairwise interactions through attention, but the associated quadratic computational cost limits effective integration of temporal dependencies across long fMRI sequences. Selective state-space models (SSMs) instead model long-range temporal dependencies implicitly through latent state evolution in a dynamical system, enabling efficient propagation of dependencies over time. However, recent SSM-based approaches for fMRI commonly operate on derived functional connectivity representations and employ single-scale temporal processing. These design choices constrain the ability to jointly represent fast transient dynamics and slower global trends within a single model. We propose NeuroSSM, a selective state-space architecture designed for end-to-end analysis of raw BOLD signals in fMRI time series. NeuroSSM addresses the above limitations through two complementary design components: a multiscale state-space backbone that captures fast and slow dynamics concurrently, and a parallel differencing branch that increases sensitivity to transient state changes. Experiments on clinical and non-clinical datasets demonstrate that NeuroSSM achieves competitive performance and efficiency against state-of-the-art fMRI analysis methods.