Activated region fitting: a robust high-power method for fMRI analysis using parameterized regions of activation

Area of Science:

• Neuroimaging
• Statistical analysis
• Brain activity mapping

Background:

• Functional MRI (fMRI) analysis often struggles with the spatial smoothness of activated brain regions.
• Current methods may lack power and interpretability when assessing activation.
• Accounting for spatial characteristics is crucial for accurate fMRI interpretation.

Purpose of the Study:

• To propose a novel method for fMRI analysis that models activated regions using Gaussian shapes.
• To perform hypothesis tests on the characteristics (location, extent, amplitude) of these modeled regions.
• To improve statistical power and ease of interpretation in fMRI data analysis.

Main Methods:

• Modeling activated brain regions with Gaussian shapes.
• Conducting hypothesis tests on the location, spatial extent, and amplitude of these Gaussian-modeled regions.
• Utilizing simulation studies and real single-subject fMRI data for validation.

Main Results:

• The proposed method demonstrates robust hypothesis testing even with model misspecification.
• Simulation studies show increased statistical power compared to standard techniques, particularly at low signal-to-noise ratios.
• Application to real fMRI data confirms enhanced power over conventional methods.

Conclusions:

• Modeling activated regions with Gaussian shapes offers a powerful and interpretable alternative for fMRI analysis.
• This approach effectively addresses the challenge of spatial smoothness in fMRI data.
• The method shows promise for improving the detection and characterization of brain activity.