Building bridges between brain and behavior: An open-source toolbox for joint modeling with fMRI

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Psychometrics

Background:

• Relating neural activity to behavior is a key challenge in cognitive neuroscience.
• Joint modeling simultaneously analyzes behavioral and fMRI data, accounting for errors and variability.
• Existing methods can struggle with parameter stability and bias in brain-behavior correlations.

Purpose of the Study:

• To present EMC2, an R package simplifying joint model estimation for behavioral and neural data.
• To demonstrate the application of EMC2 using a perceptual decision-making task in an fMRI setting.
• To provide researchers with tools for robust brain-behavior link analysis.

Main Methods:

• Developed an R package, EMC2, for joint modeling.
• Utilized a hierarchical modeling framework for stable parameter estimation.
• Applied the toolbox to fMRI and behavioral data from a perceptual decision-making task.

Main Results:

• EMC2 streamlines the estimation of behavioral, neural, and joint models.
• Joint estimation stabilizes parameters and reduces attenuation bias in brain-behavior correlations.
• The toolbox facilitates group-level and individual-difference analyses.

Conclusions:

• EMC2 offers an accessible introduction to joint modeling for researchers.
• The toolbox enables richer, more reliable insights into cognitive and neural mechanisms.
• Adoption of joint models can advance understanding of the brain-behavior relationship.