Estimation and removal of spurious echo artifacts in single-voxel MRS using sensitivity encoding

Area of Science:

• Magnetic Resonance Spectroscopy (MRS)
• Medical Imaging
• Signal Processing

Background:

• Spurious echo artifacts in localized MRS commonly overlap with metabolite resonances, hindering accurate quantification.
• These artifacts arise from unsuppressed signals outside the volume of interest.
• Sensitivity encoding offers a potential solution for separating these signal contributions.

Purpose of the Study:

• To develop and evaluate a method using sensitivity encoding to separate and remove spurious echo artifacts in localized MRS.
• To assess the effectiveness of the proposed method, termed ERASE (Echo-Artifact Removal by Sensitivity Encoding), in improving spectral quality.

Main Methods:

• Numerical simulations were conducted to explore sensitivity-encoded separation for artifact regions.
• An imaging-based approach was developed to identify potential artifact-contributing regions.
• The ERASE method was tested on six datasets from three brain regions and compared to standard coil combination.

Main Results:

• ERASE successfully unfolded artifact signals when artifact regions were known a priori.
• While mismatches reduced efficiency, metabolite signals remained unaffected.
• Water suppression imaging identified unsuppressed signal regions, and ERASE showed visible artifact removal compared to standard reconstruction, with lower fitting errors for metabolites.

Conclusions:

• The ERASE reconstruction tool effectively reduces spurious echo artifacts in single-voxel MRS.
• ERASE can be integrated into standard workflows to enhance spectral quality, particularly when hardware limitations cause out-of-voxel signal contamination.