Synthetic Data in MR Spectroscopy: Current Practices, Applications, and Considerations
Authors:
John T. LaMaster,
Aaron T. Gudmundson,
Alireza Abaei,
Seyma Alcicek,
Arturo Alvarado,
Ovidiu Andronesi,
Tiffany K. Bell,
Wolfgang Bogner,
Hanna Bugler,
Alexander R Craven,
Cristina Cudalbu,
Alma Davidson,
Christopher W. Davies-Jenkins,
Dinesh Deelchand,
Richard A. E. Edden,
Morteza Esmaeili,
Candace C Fleischer,
Abdelrahman Gad,
Guglielmo Genovese,
Saumya Gurbani,
Ashley D. Harris,
Pierre-Gilles Henry,
Kay Chioma Igwe,
Ajin Joy,
Margarida Julià-Sapé
, et al. (53 additional authors not shown)
Abstract:
The use of synthetic data has emerged as an essential tool in Magnetic Resonance Spectroscopy (MRS) research and applications, providing advantages for optimization of acquisition, software validation, deep learning applications, and enhanced reproducibility. Importantly, synthetic data addresses challenges of limited training data availability, particularly for clinical populations, and offers co…
▽ More
The use of synthetic data has emerged as an essential tool in Magnetic Resonance Spectroscopy (MRS) research and applications, providing advantages for optimization of acquisition, software validation, deep learning applications, and enhanced reproducibility. Importantly, synthetic data addresses challenges of limited training data availability, particularly for clinical populations, and offers controlled solutions for investigating uncertainties and unexplained variance with in vivo data. This work provides a review and evaluation of current practices in the use and generation of synthetic data within the MRS field. Conducted by the MRS Synthetic Data Working Group under the Code & Data Sharing Committee of the MRS Study Group of the International Society for Magnetic Resonance in Medicine (ISMRM), this manuscript encompasses existing literature, supplemented by collective experience and in-house methodologies.
△ Less
Submitted 9 March, 2026; v1 submitted 26 February, 2026;
originally announced February 2026.
A Digital Phantom for MR Spectroscopy Data Simulation
Authors:
D. M. J. van de Sande,
A. T. Gudmundson,
S. Murali-Manohar,
C. W. Davies-Jenkins,
D. Simicic,
G. Simegn,
İ. Özdemir,
S. Amirrajab,
J. P. Merkofer,
H. J. Zöllner,
G. Oeltzschner,
R. A. E. Edden
Abstract:
Simulated data is increasingly valued by researchers for validating MRS processing and analysis algorithms. However, there is no consensus on the optimal approaches for simulation models and parameters. This study introduces a novel MRS digital brain phantom framework, providing a comprehensive and modular foundation for MRS data simulation. The framework generates a digital brain phantom by combi…
▽ More
Simulated data is increasingly valued by researchers for validating MRS processing and analysis algorithms. However, there is no consensus on the optimal approaches for simulation models and parameters. This study introduces a novel MRS digital brain phantom framework, providing a comprehensive and modular foundation for MRS data simulation. The framework generates a digital brain phantom by combining anatomical and tissue label information with metabolite data from the literature. This phantom contains all necessary information for simulating spectral data. The MRS phantom is combined with a signal-based model to demonstrate its functionality and usability in generating various spectral datasets. Outputs can be saved in the NIfTI-MRS format, enabling their use in downstream applications. To evaluate the realism of the simulated spectra, a comparison was performed against in-vivo MRS data acquired under similar conditions. The phantom was implemented using two anatomical templates at different resolutions and tested across a range of user-defined simulation parameters. Simulated spectra exhibited realistic signal characteristics and structural variability. When compared to in-vivo data, the simulated spectra closely matched in terms of spectral shape, signal-to-noise ratio, and metabolite quantification. The simulations also captured key variability features and provided additional diversity not present in the in-vivo dataset, supporting use in robustness testing and data augmentation. This novel digital phantom provides a flexible and extensible platform for MRS data simulation. Its modular architecture, user-friendly GUI, and open-source implementation support reproducible research, algorithm development, and validation in the MRS community.
△ Less
Submitted 10 July, 2025; v1 submitted 20 December, 2024;
originally announced December 2024.