7T-fMRI: Faster temporal resolution yields optimal BOLD sensitivity for functional network imaging specifically at...

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Signal Processing

Background:

• Accelerated imaging methods enable faster acquisition of high-resolution images, crucial for 7 Tesla functional magnetic resonance imaging (7T-fMRI) applications like neurosurgical planning and Brain Computer Interfaces (BCIs).
• Increasing spatial and temporal resolution in fMRI can lead to signal-to-noise ratio (SNR) losses, potentially negating gains from improved temporal resolution.

Purpose of the Study:

• To investigate the impact of varying spatial and temporal resolutions on fMRI sensitivity metrics.
• To assess the implications of these resolution changes on fMRI-based BCI simulations.

Main Methods:

• Compared temporal signal-to-noise ratio (tSNR), percent signal change (%∆S), activation volume, Z-scores, and decoding performance across various spatial and temporal resolutions.
• Utilized data from an ankle-tapping task during 7T-fMRI acquisition.
• Employed linear classifiers for BCI decoding performance assessment.

Main Results:

• A 22% increase in %∆S and a 9% improvement in decoding performance were observed with higher temporal resolution at the highest spatial resolution (1.5×1.5×1.5mm³), despite a 29% decrease in tSNR.
• Z-scores plateaued, and activation volume remained consistent across spatial resolutions at the highest temporal resolution (500ms).

Conclusions:

• Overall BOLD sensitivity can be significantly enhanced by increasing temporal resolution, provided adequate spatial resolution and minimal physiological noise.
• Demonstrates the feasibility of diffuse motor-network imaging with high spatial and temporal resolution and robust BOLD sensitivity using 7T-fMRI.
• Highlights the potential for improved sensitivity in fMRI applications like BCIs.