Voxel-Based State Space Modeling Recovers Task-Related Cognitive States in Naturalistic fMRI Experiments.

Tianjiao Zhang,Tolga Çukur,Jack L Gallant,James S Gao

doi:10.3389/fnins.2020.565976

Tianjiao Zhang, Tolga Çukur + Show 2 more

Open Access

https://doi.org/10.3389/fnins.2020.565976

Copy DOI

Abstract

Complex natural tasks likely recruit many different functional brain networks, but it is difficult to predict how such tasks will be represented across cortical areas and networks. Previous electrophysiology studies suggest that task variables are represented in a low-dimensional subspace within the activity space of neural populations. Here we develop a voxel-based state space modeling method for recovering task-related state spaces from human fMRI data. We apply this method to data acquired in a controlled visual attention task and a video game task. We find that each task induces distinct brain states that can be embedded in a low-dimensional state space that reflects task parameters, and that attention increases state separation in the task-related subspace. Our results demonstrate that the state space framework offers a powerful approach for modeling human brain activity elicited by complex natural tasks.

Highlights

To maximize efficiency and statistical power, most neuroimaging experiments use simple parametric designs and highly focused data analysis
To understand how task information may be represented in the population activity space of the cortex, we developed a voxel-based state space framework for fMRI
The task-related state space is recovered by regressing task variables directly on to cortical activity, and activity at each TR can be projected to a point in this task-related state space

Summary

Introduction

To maximize efficiency and statistical power, most neuroimaging experiments use simple parametric designs and highly focused data analysis. C. et al, 2006; Krakauer et al, 2017; Yarkoni and Westfall, 2017; Matusz et al, 2019) To address this problem, some neuroimaging experiments use more complex naturalistic conditions, such as watching movies (Hasson et al, 2004; Nishimoto and Gallant, 2011; Huth et al, 2012), listening to stories (Huth et al, 2016), or playing video games (Mathiak and Weber, 2006; Spiers and Maguire, 2007; Mathiak et al, 2011). Neuroimaging data collected under naturalistic conditions elicit complex, dynamic patterns of brain activity across multiple functional networks that reflect the explicit and implicit task structure of the experiment (Çukur et al, 2013b). While watching movies selective attention to one target or another may change the representation of information in relevant functional networks (Çukur et al, 2013b), and while playing a video game the dynamic evolution of goals and subgoals over the course of the game might evoke activity in distinct functional networks over time

Objectives

Methods

Results

Conclusion