CausalMamba: Scalable Conditional State Space Models for Neural Causal Inference
This provides neuroscientists with a scalable tool for large-scale causal inference to understand cognitive function, though it is incremental as it builds on existing methods like DCM.
The paper tackled the problem of inferring neural causality from fMRI data by addressing ill-posedness and computational intractability, achieving 37% higher accuracy than Dynamic Causal Modeling on simulated data and 88% fidelity in recovering known neural pathways on real data.
We introduce CausalMamba, a scalable framework that addresses fundamental limitations in fMRI-based causal inference: the ill-posed nature of inferring neural causality from hemodynamically distorted BOLD signals and the computational intractability of existing methods like Dynamic Causal Modeling (DCM). Our approach decomposes this complex inverse problem into two tractable stages: BOLD deconvolution to recover latent neural activity, followed by causal graph inference using a novel Conditional Mamba architecture. On simulated data, CausalMamba achieves 37% higher accuracy than DCM. Critically, when applied to real task fMRI data, our method recovers well-established neural pathways with 88% fidelity, whereas conventional approaches fail to identify these canonical circuits in over 99% of subjects. Furthermore, our network analysis of working memory data reveals that the brain strategically shifts its primary causal hub-recruiting executive or salience networks depending on the stimulus-a sophisticated reconfiguration that remains undetected by traditional methods. This work provides neuroscientists with a practical tool for large-scale causal inference that captures both fundamental circuit motifs and flexible network dynamics underlying cognitive function.