T2 relaxometry with indirect echo compensation from highly undersampled data.

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Physics
• Biomedical Engineering

Background:

• Accurate T2 estimation is crucial for various MRI applications.
• Undersampled multi-echo spin-echo data presents challenges for precise T2 mapping.
• Indirect echoes can introduce significant biases in T2 quantification.

Purpose of the Study:

• To develop a fast and accurate algorithm for T2 estimation.
• To address challenges posed by highly undersampled multi-echo spin-echo data.
• To minimize T2 biases arising from indirect echoes.

Main Methods:

• A model-based reconstruction approach using the slice-resolved extended phase graph (SEPG) model.
• Principal component decomposition to linearize the nonlinear SEPG signal model.
• The CUrve Reconstruction via principal component-based Linearization with Indirect Echo compensation (CURLIE) algorithm for T2 curve estimation.

Main Results:

• Phantom studies showed significant T2 biases (< 3.2%) when indirect echoes were accounted for using CURLIE and SEPG fitting, even at 4% sampling.
• Uncorrected indirect echoes led to T2 biases ranging from 1.9% to 18.4% in phantom experiments.
• In vivo experiments in brain, liver, and heart demonstrated similar T2 estimation trends as observed in phantoms.

Conclusions:

• The CURLIE reconstruction combined with SEPG fitting provides accurate T2 estimation from highly undersampled radial multi-echo spin-echo data.
• This method offers a fast T2 mapping solution.
• It effectively eliminates errors caused by indirect echoes, enhancing diagnostic reliability.