T1-weighting in Steady-State FLASH MRI-Diffusion Is Not Only Supportive but Mandatory for the Contrast

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Physics
• Computational Modeling

Background:

• Fast Low Angle Shot (FLASH) imaging is commonly used for MRI.
• It is assumed to produce T1-weighted steady-state contrast through RF and gradient spoiling.
• Previous simulations showed overestimation of CSF signals when diffusion was ignored.

Purpose of the Study:

• To investigate the role of diffusion in steady-state FLASH contrast formation.
• To assess the implications of diffusion on simulation-based modeling and measurements in MRI.

Main Methods:

• Simulated FLASH sequences using phase graph simulations with a synthetic brain phantom.
• Evaluated the impact of neglecting diffusion on a segmentation network trained on simulated data.
• Validated contrast changes experimentally using a 3D-printed phantom with silicone oil.

Main Results:

• Simulations without diffusion showed higher CSF signal intensities than white matter (WM).
• Diffusion was found to suppress higher-order echoes in long T2 tissues, essential for T1-weighted contrast.
• A neural network trained without diffusion failed to generalize to in vivo data.

Conclusions:

• Diffusion is essential for realistic FLASH simulations of tissues with long T2 relaxation times, like CSF.
• Steady-state FLASH contrast results from an interplay of RF spoiling, gradient spoiling, and relaxation-spoiling influenced by T2 decay and diffusion.
• Diffusion must be included in MRI simulations and simulation-based deep learning applications for accurate T1-weighted contrast.