FSL-MRS: An end-to-end spectroscopy analysis package

Area of Science:

• Neuroimaging
• Spectroscopy
• Medical Physics

Background:

• Magnetic Resonance Spectroscopy (MRS) is crucial for non-invasive metabolite quantification.
• Existing MRS analysis tools can be limited in flexibility and modularity.
• An integrated, open-source solution is needed for advanced MRS data processing.

Purpose of the Study:

• Introduce FSL-MRS, an end-to-end, modular, open-source toolbox for MRS analysis.
• Provide tools for data conversion, preprocessing, simulation, fitting, quantitation, and visualization.
• Enable flexible analysis of various MRS acquisition types.

Main Methods:

• FSL-MRS utilizes the Neuroimaging Informatics Technology Initiative (NIfTI) format for MRS data.
• Includes preprocessing steps like coil combination and frequency/phase alignment.
• Employs density matrix simulation for basis spectra and Markov Chain Monte Carlo for metabolite quantitation.

Main Results:

• Demonstrated low fitting errors on simulated (11.9%) and phantom (3.4%) data.
• Achieved high correlation (0.53-0.81) with LCModel on in vivo human datasets.
• Validated across single-voxel and multivoxel spectroscopy at 3 T and 7 T.

Conclusions:

• FSL-MRS provides a flexible and extensible platform for MRS analysis.
• Custom fitting models can be integrated for advanced applications.
• The toolbox is available as part of the FMRIB Software Library.