A mixed-effects, spatially varying coefficients model with application to multi-resolution functional magnetic...

Area of Science:

• Neuroimaging
• Medical Physics
• Radiology

Background:

• Functional magnetic resonance imaging (fMRI) is crucial for pre-surgical planning, balancing spatial resolution and signal-to-noise ratio (SNR).
• Standard fMRI offers high SNR but lower spatial accuracy, while high-resolution fMRI provides greater spatial detail at the cost of reduced SNR.
• Optimizing SNR and spatial accuracy in pre-operative fMRI is essential for precise surgical guidance.

Purpose of the Study:

• To develop and evaluate a novel method for integrating standard and high-resolution fMRI data for pre-surgical analysis.
• To assess if combining the high SNR of standard fMRI with the spatial accuracy of high-resolution fMRI can improve pre-operative functional mapping.
• To compare the proposed model's performance against existing single-resolution approaches.

Main Methods:

• A mixed-effects model with spatially varying coefficients was proposed to regress high-resolution fMRI statistic images onto standard-resolution fMRI statistic images.
• The model's efficacy was evaluated through comprehensive simulation studies.
• Performance was benchmarked against a recently developed single-resolution model using both simulated and real patient data.

Main Results:

• Simulation studies demonstrated the proposed model's ability to effectively combine SNR from standard resolution and spatial accuracy from high resolution scans.
• Real-data analysis on a patient awaiting tumor resection confirmed the model's practical utility.
• The newly proposed model outperformed the single-resolution model in leveraging both SNR and spatial resolution advantages.

Conclusions:

• The developed model successfully integrates standard and high-resolution fMRI data, offering improved pre-surgical functional mapping.
• This approach enhances SNR from standard scans while preserving the spatial accuracy of high-resolution scans.
• The findings suggest a significant advancement in pre-operative fMRI analysis for neurosurgical applications.