Aliasing-free reduced field-of-view parallel imaging.

Area of Science:

• Magnetic Resonance Imaging
• Image Reconstruction

Background:

• Reduced Field-of-View (rFOV) parallel imaging (PI) methods often introduce aliasing artifacts.
• Existing PI techniques struggle to reconstruct aliasing-free full FOV images from rFOV data, particularly with Cartesian and Wave sampling.

Purpose of the Study:

• To develop a method for reconstructing aliasing-free full FOV images from rFOV PI data.
• To address aliasing artifacts inherent in rFOV PI with Cartesian and Wave sampling.

Main Methods:

• Extended the Sensitivity Encoding (SENSE) method to Soft-SENSE models.
• Utilized multiple-set coil sensitivity maps (CSM) and point spread functions (PSF) derived from full FOV data for rFOV PI.
• Applied Soft-SENSE reconstruction algorithms compatible with conventional full FOV SENSE.

Main Results:

• Successfully reconstructed aliasing-free full FOV images from rFOV PI data using Soft-SENSE with Cartesian and Wave sampling.
• Demonstrated flexible aliasing control and comparable geometry factors to standard full FOV PI.
• Showcased improved computational efficiency for iterative compressed sensing (CS) and PI reconstruction due to reduced Fast Fourier Transform (FFT) size.

Conclusions:

• Soft-SENSE with full FOV CSM and PSF is a viable solution for rFOV PI.
• Enables more flexible PI acceleration and enhances computational efficiency for iterative CSPI reconstruction.