FAST-Brain: A Flow-Aligned Spatio-Temporal Surrogate Brain Model

2026-09-28

AI Summary

This study addresses the limitations of conventional methods in capturing long-range temporal dynamics, anatomical spatial structures, and the low-dimensional nature of high-dimensional brain signals. We propose a unified flow-aligned spatiotemporal surrogate brain model that integrates graph convolutional networks with Transformer architectures to directly predict BOLD signals. Theoretically, we demonstrate that under a low-dimensional subspace assumption, the approximation error scales solely with the intrinsic dimensionality rather than the ambient dimensionality, thereby providing rigorous theoretical justification for direct modeling. Empirical evaluations on both synthetic and Human Connectome Project (HCP) datasets reveal that the proposed model achieves state-of-the-art performance in functional connectivity estimation, effective connectivity recovery, and implicit low-dimensional subspace representation.

Abstract

Modeling resting-state functional magnetic resonance imaging (rs-fMRI) data is crucial for understanding brain-wide neural activity. However, traditional methods struggle to capture complex temporal dynamics over long horizons, to account for the brain's anatomical spatial structure, and to model high-dimensional ambient signals that lie on a low-dimensional intrinsic subspace. We propose FAST-Brain, a unified flow-aligned spatio-temporal surrogate brain model that addresses all three challenges. At its core is a flow-aligned generative framework that directly predicts the clean blood-oxygen-level-dependent (BOLD) signal, paired with a graph convolutional network that captures spatial structural constraints and a Transformer that models long-range temporal dependencies. Theoretically, we show that under a low-dimensional subspace assumption, the approximation error of our model scales with the intrinsic dimension rather than the ambient dimension, which justifies our direct modeling of the BOLD signal. Extensive experiments on synthetic and Human Connectome Project datasets demonstrate that FAST-Brain achieves state-of-the-art performance in recovering functional connectivity, effective connectivity, and the implicit low-dimensional signal subspace.

Problem

• rs-fMRI
• temporal dynamics
• spatial structure
• low-dimensional subspace
• surrogate brain model

Innovation

• Flow-Aligned Generative Framework
• Spatio-Temporal Surrogate Model
• Graph Convolutional Network
• Transformer
• Low-Dimensional Subspace