Fast T2 mapping with multiple echo, Caesar cipher acquisition and model-based reconstruction.

Area of Science:

• Magnetic Resonance Imaging
• Quantitative Imaging
• Biomedical Engineering

Background:

• Fast, quantitative T2 mapping is crucial for clinical and research applications.
• Accelerated Fast Spin Echo (FSE) sequences often suffer from artifacts, impacting parameter map accuracy.
• Imperfect refocusing further compounds these errors in T2 estimation.

Purpose of the Study:

• To present a B1-corrected, model-based reconstruction method for enhanced accuracy in accelerated T2 mapping.
• To introduce a Cartesian FSE phase-encode ordering to improve T2 estimates.
• To overcome limitations of existing accelerated T2 mapping protocols.

Main Methods:

• Developed Multiple Echo, Caesar Cipher Acquisition and Model-Based Reconstruction (ME-CAMBREC).
• Directly fitted T2, flip angle, and proton density maps using the extended phase graph (EPG) model on k-space data.
• Employed regularization to mitigate noise amplification and evaluated in phantoms and human brain.

Main Results:

• ME-CAMBREC demonstrated fewer artifacts and comparable or lower error rates than other methods at moderate-to-high acceleration.
• In vivo studies showed ME-CAMBREC achieved approximately half the error rates of alternative FSE-based T2 mapping protocols.
• The method proved effective in computational, phantom, and human brain experiments.

Conclusions:

• Directly fitting multi-echo data to k-space using the extended phase graph model significantly enhances T2 map fidelity.
• Appropriate phase-encode ordering is critical for maximizing the benefits of this model-based approach.
• ME-CAMBREC offers a more accurate solution for accelerated quantitative T2 mapping.