Cyclic Continuous Max-Flow: A Third Paradigm in Generating Local Phase Shift Maps in MRI.

Area of Science:

• Medical Imaging
• Image Processing
• Magnetic Resonance Imaging (MRI)

Background:

• MRI phase deviation sensitivity underpins techniques like magnetic susceptibility weighted imaging and chemical shift imaging.
• Existing phase processing methods, either coupled or unwrapping-based, are prone to errors from low signal and phase poles.

Purpose of the Study:

• To introduce Cyclic Continuous Max-Flow (CCMF) as a new paradigm for MRI phase processing.
• To enhance the robustness and accuracy of MRI phase processing, overcoming limitations of current techniques.

Main Methods:

• Developed CCMF phase processing utilizing primal-dual-variational optimization.
• Implemented CCMF over a cylindrical manifold to represent inherent phase image topology.
• Compared CCMF performance against existing coupled and unwrapping-based phase processing methods.

Main Results:

• CCMF demonstrates increased robustness to issues like low signal and phase poles.
• The technique maintains accuracy comparable to existing methods.
• CCMF proves to be an efficient and robust phase processing solution.

Conclusions:

• CCMF represents a distinct and advanced approach to MRI phase processing.
• The inherent topological representation in CCMF significantly improves error resilience.
• CCMF offers a promising alternative for accurate and reliable MRI phase analysis.