Multiband phase-constrained parallel MRI.

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Physics
• Biomedical Engineering

Background:

• Parallel MRI methods often reduce signal-to-noise ratio (SNR), limiting acceleration.
• Multiband (MB) radio-frequency (RF) pulses excite multiple slices simultaneously to boost intrinsic SNR.
• Separating closely spaced slices in MB imaging can amplify noise due to insufficient coil sensitivity variation.

Purpose of the Study:

• To apply phase-constrained reconstruction for multiband MRI experiments.
• To minimize noise amplification during slice separation in parallel MRI.
• To improve SNR performance in accelerated multi-slice imaging.

Main Methods:

• Induced pre-defined phase differences between neighboring slices.
• Utilized phase-constrained parallel MRI reconstruction for slice separation.
• Tailored inter-slice phase differences for optimal separation and noise reduction.
• Demonstrated the method with in-vivo multiband experiments.

Main Results:

• Significantly reduced noise amplification in phase-constrained multiband reconstructions compared to standard methods.
• Optimal noise reduction achieved with a 90° phase difference between neighboring slices (12 mm apart).

Conclusions:

• Phase-constrained parallel MRI offers a viable approach for accelerated multi-slice imaging.
• The method demonstrates improved SNR performance in multiband acquisitions.
• This technique holds potential for enhancing diagnostic quality in rapid MRI scans.