Hybrid-Space SENSE Reconstruction for Simultaneous Multi-Slice MRI.

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Imaging Physics
• Image Reconstruction

Background:

• Simultaneous Multi-Slice (SMS) MRI accelerates imaging by using controlled aliasing through periodic undersampling.
• Evaluating undersampling patterns requires quantitative SNR loss assessment.
• Echo Planar Imaging (EPI) in SMS MRI suffers from slice-specific Nyquist ghosting artifacts due to eddy currents, challenging accurate correction.

Purpose of the Study:

• To develop a hybrid-space SENSE reconstruction framework for SMS MRI using a 3D representation.
• To derive analytical maps for quantifying SNR loss in SMS acquisitions with arbitrary undersampling patterns.
• To introduce a matrix-decoding method for correcting slice-specific Nyquist ghosting in SMS EPI.

Main Methods:

• Proposed a hybrid-space SENSE reconstruction framework for SMS MRI.
• Derived analytical Signal-to-Noise Ratio (SNR) loss maps for various undersampling patterns.
• Developed a matrix-decoding correction technique for slice-specific Nyquist ghosting in SMS EPI.

Main Results:

• The hybrid-space SENSE reconstruction produced images comparable in quality to existing methods.
• Analytical SNR loss maps closely matched Monte Carlo results but were computationally faster.
• The matrix-decoding method demonstrated superior performance in correcting Nyquist ghosting compared to single-slice and slice-averaged methods.

Conclusions:

• The proposed hybrid-space SENSE framework offers an effective approach for SMS MRI reconstruction.
• Analytical SNR loss mapping provides an efficient tool for comparing SMS undersampling strategies.
• The matrix-decoding method significantly improves artifact correction in SMS EPI, enhancing image quality.