Compressed sensing fMRI using gradient-recalled echo and EPI sequences.

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Signal Processing

Background:

• Functional magnetic resonance imaging (fMRI) is crucial for neuroscience research.
• Accelerating fMRI data acquisition is desirable for high-resolution imaging.
• Compressed sensing (CS) offers potential for fMRI acceleration but requires further investigation due to signal dynamics and power loss concerns.

Purpose of the Study:

• To systematically evaluate the utility of compressed sensing for functional MRI (CS-fMRI).
• To investigate the impact of CS on functional sensitivity, specificity, and temporal characteristics in fMRI.
• To compare CS-fMRI performance with fully sampled data using different imaging sequences and undersampling factors.

Main Methods:

• Computer simulations and in vivo experiments were conducted using rat sensory and odor stimulation paradigms.
• Gradient-recalled echo (GRE) and echo planar imaging (EPI) sequences were employed.
• k-t FOCUSS algorithm was used for CS reconstruction with varying undersampling patterns and reduction factors (2 and 4).

Main Results:

• CS-fMRI demonstrated improved statistical sensitivity for activation detection with GRE sequences, especially under low signal-to-noise conditions.
• CS enhanced temporal resolution and reduced temporal noise correlations.
• While CS reduced functional response amplitudes, the decreased noise variance led to more sensitive activation detection.

Conclusions:

• Compressed sensing is a valuable approach for accelerating fMRI data acquisition.
• CS-fMRI shows particular promise for gradient-recalled echo fMRI studies, enhancing sensitivity without significant statistical power loss.
• The findings support the broader application of CS techniques in fMRI research.