Brain dynamics and spatiotemporal trajectories during threat processing.

In the past decades, functional MRI research has investigated mental states and their brain bases in largely static fashion based on evoked responses during blocked and event-related designs. Despite some progress in naturalistic designs, our understanding of threat processing remains largely limited to those obtained with standard paradigms. In the present paper, we applied Switching Linear Dynamical Systems to uncover the dynamics of threat processing during a continuous threat-of-shock paradigm. Importantly, unlike studies in systems neuroscience that frequently assume that systems are decoupled from external inputs, we characterized both endogenous and exogenous contributions to dynamics. First, we demonstrated that the SLDS model learned the regularities of the experimental paradigm, such that states and state transitions estimated from fMRI time series data from 85 ROIs reflected both the proximity of the circles and their direction (approach vs. retreat). After establishing that the model captured key properties of threat-related processing, we characterized the dynamics of the states and their transitions. The results revealed that threat processing can profitably be viewed in terms of dynamic multivariate patterns whose trajectories are a combination of intrinsic and extrinsic factors that jointly determine how the brain temporally evolves during dynamic threat. We propose that viewing threat processing through the lens of dynamical systems offers important avenues to uncover properties of the dynamics of threat that are not unveiled with standard experimental designs and analyses.

Learning brain dynamics for decoding and predicting individual differences.

Insights from functional Magnetic Resonance Imaging (fMRI), as well as recordings of large numbers of neurons, reveal that many cognitive, emotional, and motor functions depend on the multivariate interactions of brain signals. To decode brain dynamics, we propose an architecture based on recurrent neural networks to uncover distributed spatiotemporal signatures. We demonstrate the potential of the approach using human fMRI data during movie-watching data and a continuous experimental paradigm. The model was able to learn spatiotemporal patterns that supported 15-way movie-clip classification (∼90%) at the level of brain regions, and binary classification of experimental conditions (∼60%) at the level of voxels. The model was also able to learn individual differences in measures of fluid intelligence and verbal IQ at levels comparable to that of existing techniques. We propose a dimensionality reduction approach that uncovers low-dimensional trajectories and captures essential informational (i.e., classification related) properties of brain dynamics. Finally, saliency maps and lesion analysis were employed to characterize brain-region/voxel importance, and uncovered how dynamic but consistent changes in fMRI activation influenced decoding performance. When applied at the level of voxels, our framework implements a dynamic version of multivariate pattern analysis. Our approach provides a framework for visualizing, analyzing, and discovering dynamic spatially distributed brain representations during naturalistic conditions.

Resting state dynamics meets anatomical structure: Temporal multiple kernel learning (tMKL) model.

Over the last decade there has been growing interest in understanding the brain activity, in the absence of any task or stimulus, captured by the resting-state functional magnetic resonance imaging (rsfMRI). The resting state patterns have been observed to be exhibiting complex spatio-temporal dynamics and substantial effort has been made to characterize the dynamic functional connectivity (dFC) configurations. However, the dynamics governing the state transitions that the brain undergoes and their relationship to stationary functional connectivity still remains an open problem. One class of approaches attempts to characterize the dynamics in terms of finite number of latent brain states, however, such attempts are yet to amalgamate the underlying anatomical structural connectivity (SC) with the dynamics. Another class of methods links individual dynamic FCs with the underlying SC but does not characterize the temporal evolution of FC. Further, the latent states discovered by previous approaches could not be directly linked to the SC, thereby motivating us to discover the underlying lower-dimensional manifold that represents the temporal structure. In the proposed approach, the discovered manifold is further parameterized as a set of local density distributions, or latent transient states. We propose an innovative method that learns parameters specific to the latent states using a graph-theoretic model (temporal Multiple Kernel Learning, tMKL) that inherently links dynamics to the structure and finally predicts the grand average FC of the test subjects by leveraging a state transition Markov model. The proposed solution does not make strong assumptions about the underlying data and is generally applicable to resting or task data for learning subject-specific state transitions and for successfully characterizing SC-dFC-FC relationship through a unifying framework. Training and testing were done using the rs-fMRI data of 46 healthy participants. tMKL model performs significantly better than the existing models for predicting resting state functional connectivity based on whole-brain dynamic mean-field model (DMF), single diffusion kernel (SDK) model and multiple kernel learning (MKL) model. Further, the learned model was tested on an independent cohort of 100 young, healthy participants from the Human Connectome Project (HCP) and the results establish the generalizability of the proposed solution. More importantly, the model retains sensitivity toward subject-specific anatomy, a unique contribution towards a holistic approach for SC-FC characterization.