Data-driven Synthesis of Magnetic Resonance Spectroscopy Data using a Variational Autoencoder

2026-02-28

AI Summary

This study addresses the scarcity of high-quality in vivo magnetic resonance spectroscopy (MRS) data for deep learning by proposing the first purely data-driven framework for single-voxel MRS generation. Built upon a complex-valued variational autoencoder (VAE), the method learns a low-dimensional latent representation of real spectra and synthesizes new samples through latent space sampling and interpolation, bypassing reliance on imperfect physical simulation models. The generated spectra accurately reproduce dominant spectral patterns and share a common feature space with real data. When applied to GABA-edited MRS, synthetic data significantly improve signal-to-noise ratio, linewidth, and spectral quality ratings, though biases persist in absolute metabolite quantification and noise modeling remains suboptimal. The work further introduces an application-aware evaluation framework to systematically assess the utility and limitations of generated data for downstream tasks.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

The development of deep learning methods for magnetic resonance spectroscopy (MRS) is often hindered by limited availability of large, high-quality training datasets. While physics-based simulations are commonly used to mitigate this limitation, accurately modeling all in-vivo signal components remains challenging. In this work, we propose a data-driven framework for synthesizing in-vivo MRS data using a variational autoencoder (VAE) trained exclusively on measured single-voxel spectroscopy data. The model learns a low-dimensional latent representation of complex-valued spectra and enables generation of new samples through latent-space sampling and interpolation. The generative performance of the proposed approach is evaluated using a comprehensive set of complementary analyses, including reconstruction quality, feature-level similarity using low-dimensional embeddings, application-based signal quality metrics, and metabolite quantification agreement. The results demonstrate that the VAE accurately reconstructs dominant spectral patterns and generates synthetic spectra that occupy the same feature space as in-vivo data. In an example application targeting GABA-edited spectroscopy, augmenting limited subsets of transients with synthetic spectra improves signal quality metrics such as signal-to-noise ratio, linewidth, and shape scores. However, the results also reveal limitations of the generative approach, including under-representation of stochastic noise and reduced accuracy in absolute metabolite quantification, particularly for applications sensitive to concentration estimates. These findings highlight both potential and limitations of data-driven MRS synthesis. Beyond the proposed model, this study introduces a structured evaluation framework for generative MRS methods, emphasizing the importance of application-aware validation when synthetic data are used for downstream analysis.

Problem

• Magnetic Resonance Spectroscopy
• data synthesis
• data scarcity
• deep learning
• in-vivo MRS

Innovation

• variational autoencoder
• magnetic resonance spectroscopy
• data-driven synthesis
• latent space sampling
• generative model evaluation