A hierarchical model for simultaneous detection and estimation in multi-subject fMRI studies.

Area of Science:

• Neuroimaging
• Statistical Modeling
• Cognitive Neuroscience

Background:

• Standard multi-subject fMRI analysis faces challenges in modeling individual hemodynamic response variations.
• Accurate estimation of population-level brain activation requires accounting for regional and subject-specific response shapes.

Purpose of the Study:

• To introduce a new hierarchical model for simultaneous brain activation detection and hemodynamic response function (HRF) shape estimation in multi-subject fMRI.
• To overcome limitations of standard fMRI analysis by allowing flexible HRF shapes across regions and subjects.
• To provide a method for estimating population-level activation while accommodating individual variability.

Main Methods:

• Development of a novel hierarchical statistical model for fMRI data.
• Implementation of an efficient estimation algorithm for model parameters.
• Creation of an inferential framework for testing brain activation and HRF shape deviations.

Main Results:

• The model successfully allows the HRF shape to vary across regions and subjects.
• It provides a straightforward method for estimating population-level activation.
• Validation through simulations and application to a thermal pain fMRI study demonstrated model efficacy.

Conclusions:

• The proposed hierarchical model offers a flexible and robust approach to multi-subject fMRI analysis.
• It enhances the ability to detect brain activation and characterize the hemodynamic response.
• This method advances the analysis of complex brain function in group studies.