PyAMARES, an Open-Source Python Library for Fitting Magnetic Resonance Spectroscopy Data.

Area of Science:

• Metabolic research
• Biomedical engineering
• Computational biology

Background:

• Magnetic resonance spectroscopy (MRS) is crucial for in vivo metabolic studies.
• The Advanced Method for Accurate, Robust, and Efficient Spectral fitting (AMARES) is a widely used quantification technique.
• Existing AMARES implementations may lack flexibility and integration with modern computational tools.

Purpose of the Study:

• Introduce pyAMARES, an open-source Python implementation of the AMARES algorithm.
• Provide a flexible, user-friendly, and versatile tool for MRS quantification.
• Enhance reproducibility and collaboration in MRS research.

Main Methods:

• Developed pyAMARES as a Python library implementing the AMARES algorithm.
• Incorporated multiprocessing and customizable objective functions.
• Validated pyAMARES against OXSA and jMRUI using simulated and in vivo MRS data via Monte Carlo simulations.

Main Results:

• PyAMARES demonstrated comparable accuracy, precision, and computational efficiency to existing software.
• Features include spreadsheet-based settings, fitting without prior knowledge, and metabolite residual removal.
• Multiprocessing significantly improved pyAMARES performance.

Conclusions:

• PyAMARES provides a robust and flexible solution for MRS quantification in the Python ecosystem.
• Its open-source nature, documentation, and integration foster reproducibility and collaboration.
• PyAMARES bridges traditional MRS fitting with machine learning, advancing metabolic research and clinical applications.