Model-Based Chemical Exchange Saturation Transfer MRI for Robust z-Spectrum Analysis

Area of Science:

• Medical Imaging
• Biophysics
• Spectroscopy

Background:

• Chemical Exchange Saturation Transfer (CEST) Magnetic Resonance Imaging (MRI) is a sensitive technique for detecting low-concentration metabolites.
• Robust z-spectrum analysis is crucial for accurate CEST quantification, but challenges remain in handling asymmetric spectral components.

Purpose of the Study:

• To introduce a novel model-based CEST MRI method incorporating subspace-based spectral decomposition.
• To enable direct estimation of asymmetric spectra from complete or incomplete measurements for improved z-spectrum analysis.

Main Methods:

• Developed a model-based CEST MRI approach integrating subspace-based spectral signal decomposition.
• Decomposed spectral signals into symmetric and asymmetric components using constrained optimization.
• Employed linearized spectral decomposition, weighted Frobenius norm regularization, spatial sparsity, and low-rank priors.

Main Results:

• Successfully demonstrated the feasibility of the proposed method through simulations and in vivo experiments.
• Showcased robust z-spectrum analysis by accurately estimating asymmetric spectral components.
• Validated the method's potential for reliable molecular MRI.

Conclusions:

• The proposed model-based CEST MRI with subspace spectral decomposition offers a reliable approach for molecular imaging.
• This method enhances the robustness of z-spectrum analysis in CEST MRI.
• It holds promise for advancing molecular MRI applications.