Joint maximum likelihood estimation of activation and Hemodynamic Response Function for fMRI.

Area of Science:

• Neuroimaging
• Biophysics
• Signal Processing

Background:

• Blood Oxygen Level Dependent (BOLD) functional magnetic resonance imaging (fMRI) is crucial for mapping brain activity.
• Accurate brain activity mapping relies on understanding the Hemodynamic Response Function (HRF), which varies individually and regionally.
• Estimating HRF and identifying active brain voxels are interdependent challenges in fMRI analysis.

Purpose of the Study:

• To develop a joint estimation method for HRF and brain activation using fMRI data.
• To improve the accuracy of fMRI analysis by accounting for subject- and region-specific HRF variations.
• To validate the proposed method using both synthetic and real fMRI datasets.

Main Methods:

• Employs joint maximum likelihood estimation for HRF and activation.
• Utilizes low-rank matrix approximations on regions of interest (ROIs).
• Applies Tikhonov regularization for HRF smoothing due to limited ROI data.

Main Results:

• Demonstrates accurate HRF estimation from synthetic data without prior shape assumptions.
• The method performs reliably under both white and colored noise conditions.
• Successful validation using real fMRI data from auditory stimulus experiments.

Conclusions:

• The proposed joint estimation method effectively determines HRF and brain activation in fMRI.
• This technique offers a robust approach for analyzing fMRI data with improved accuracy.
• The findings support the utility of this method for advancing neuroimaging research.