Modelling Motion-Induced Signal Corruption in Steady-State Diffusion MRI

Area of Science:

• Magnetic Resonance Imaging
• Diffusion Imaging
• Biophysics

Background:

• Diffusion-weighted steady-state free precession (DW-SSFP) offers high signal-to-noise ratio (SNR) efficiency for diffusion imaging.
• Subject motion is a significant challenge in in vivo DW-SSFP, limiting its application to low-motion scenarios.
• Accurate diffusion measurements are crucial for understanding tissue microstructure and disease.

Purpose of the Study:

• To establish a framework for modeling and correcting subject motion in DW-SSFP imaging.
• To address the motion sensitivity limitations of DW-SSFP for in vivo applications.
• To enable robust diffusion tensor estimation in the presence of physiological motion.

Main Methods:

• Developed an extended phase graphs (EPG) model incorporating a motion operator for DW-SSFP signals.
• Validated the EPG-motion model using Monte Carlo simulations.
• Integrated the model into a data fitting routine for motion estimation and correction, applied to in vivo human brain data.

Main Results:

• The EPG-motion framework demonstrated excellent agreement with simulations, showing robust diffusion coefficient estimation across various motion and SNR levels.
• Motion-corrected DW-SSFP tensor estimates showed good visual agreement with diffusion-weighted spin-echo (DW-SE) data.
• The method significantly reduced orientation-dependent motion-induced biases in diffusion tensor imaging.

Conclusions:

• Temporal signal evolution in DW-SSFP can be leveraged for retrospective motion estimation and correction.
• The developed framework enables the reconstruction of motion-corrected DW-SSFP data.
• Open-source software is provided to facilitate future research on motion impacts in DW-SSFP acquisitions.