Abstract:Decoded Neurofeedback (DecNef) represents a pioneering approach in human neuroscience that enables modulation of brain activity patterns without subjective conscious awareness through the combination of real-time fMRI with multivariate pattern analysis. While this technique holds significant potential for clinical and cognitive applications, the causal mechanisms underlying successful DecNef regulation and the neural dynamics that distinguish successful learners from those who struggle remain poorly understood. To address this question, we conducted a meta-study across functional magnetic resonance imaging (fMRI) data from five DecNef experiments, each with multiple fMRI sessions, to reveal causal network dynamics associated with individual differences in neurofeedback performance. Using the newly proposed CaLLTiF causal discovery method, we computed causal maps to identify causal network patterns that distinguish DecNef regulation from baseline and account for variations in neurofeedback success. We found that enhanced connectivity within the bilateral control network--particularly stronger connections involving the posterior cingulate and precuneus cortex--predicted neurofeedback success across all five studies. Whole-brain causal connectivity during DecNef further exhibited distinct network reorganizations, characterized by reduced average path lengths and increased right-limbic nodal degrees. Further, comparisons across cognition- and perception-targeted DecNef revealed a remarkable separation in connections to and from the somatomotor network, where connections between somatomotor and control-default-attention networks are larger during cognitive neurofeedback while causal effects between somatomotor and subcortical-visual-limbic networks are larger during perceptive DecNef. This is despite the fact that none of the involved studies targeted or involved motor activity. Overall, our results demonstrated the key role of bilateral medial control network in successful DecNef regulation regardless of the DecNef targets, a clear separation in somatomotor involvement between cognitive and perceptive DecNef, and general promise of whole-brain causal discovery in understanding complex neural processes such as decoded neurofeedback.

MetaCAE: Causal autoencoder with meta-knowledge transfer for brain effective connectivity estimation

Modeling Task Fmri Data Via Deep Convolutional Autoencoder

MCAN: Multimodal Causal Adversarial Networks for Dynamic Effective Connectivity Learning from fMRI and EEG Data

Accurate estimation of functional brain connectivity via Bayesian ICA with population-derived priors

Functional connectome fingerprinting: Identifying individuals and predicting cognitive functions via autoencoder

A Deconvolution-Based Approach to Identifying Large-Scale Effective Connectivity.

Perturbing a Neural Network to Infer Effective Connectivity: Evidence from Synthetic EEG Data

BrainMAE: A Region-aware Self-supervised Learning Framework for Brain Signals

Amortization Transformer for Brain Effective Connectivity Estimation from fMRI Data

Data-Efficient Brain Connectome Analysis via Multi-Task Meta-Learning

Brain effective connectome based on fMRI and DTI data: Bayesian causal learning and assessment

Whole-brain causal connectivity during decoded neurofeedback: a meta study

Effective Connectivity-Based Neural Decoding: A Causal Interaction-Driven Approach

Investigating Dynamic Causal Network with Unified Granger Causality Analysis

A Framework for Exploring Non-Linear Functional Connectivity and Causality in the Human Brain: Mutual Connectivity Analysis (MCA) of Resting-State Functional MRI with Convergent Cross-Mapping and Non-Metric Clustering

Exploring Brain Hemodynamic Response Patterns Via Deep Recurrent Autoencoder

RS-MAE: Region-State Masked Autoencoder for Neuropsychiatric Disorder Classifications Based on Resting-State fMRI

Mapping dynamic spatial patterns of brain function with spatial-wise attention

Functional Connectivity Networks with Latent Distributions for Mild Cognitive Impairment Identification

Inferring brain causal and temporal-lag networks for recognizing abnormal patterns of dementia

Uncovering cognitive taskonomy through transfer learning in masked autoencoder-based fMRI reconstruction