Optimizing Selective RF Pulses for Enhanced Signal Stability in Turbo Spin Echo Using a Differentiable Extended Phase...

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Pulse Sequence Design
• Quantitative Imaging

Background:

• Turbo Spin Echo (TSE) imaging is crucial for various MRI applications.
• Slice profile inconsistency in TSE can lead to image blurring and inaccurate quantitative measurements, particularly for T2 mapping.
• Existing methods for RF pulse design may not adequately address slice profile consistency across echo trains.

Purpose of the Study:

• To develop and validate a novel method for optimizing radiofrequency (RF) pulses in TSE imaging.
• The primary goal is to enhance slice profile consistency throughout the echo train.
• This aims to reduce image blurring and improve the accuracy of multi-echo spin echo T2 mapping.

Main Methods:

• Utilized a differentiable extended phase graph (EPG) model incorporating RF pulse spinor profiles to calculate slice profiles.
• Employed an L-BFGS optimization algorithm with singular value regularization in PyTorch to minimize signal magnitude errors.
• Compared optimized pulses against time-bandwidth-matched Shinnar-Le Roux (SLR) RF pulses through simulations, phantom studies, and in vivo imaging.

Main Results:

• Optimized pulses achieved a 90% reduction in the standard deviation of normalized integrated signal across echoes compared to SLR pulses.
• Demonstrated increased in vivo image sharpness at tissue-fluid and vessel boundaries.
• Reduced T2 mapping error by 91% in a NIST phantom and yielded more accurate in vivo T2 maps.

Conclusions:

• The developed optimization method allows for flexible RF pulse design in echo train sequences.
• Achieved consistent slice profiles, target signal progression, and constant phase and Full Width at Half Maximum (FWHM) between echoes.
• This approach significantly improves the quality and accuracy of TSE imaging and T2 mapping.