Chemical shift and relaxation regularization improve the accuracy of ¹H MR spectroscopy analysis

Accurate analysis of metabolite levels from ¹H MRS data is a significant challenge, typically requiring the estimation of approximately 100 parameters from a single spectrum. Signal overlap, spectral noise, and common artifacts further complicate the analysis, leading to instability and reports of poor agreement between different analysis approaches. One inconsistently used method to improve analysis stability is known as regularization, where poorly determined parameters are partially constrained to take a predefined value. In this study, we examine how regularization of frequency and linewidth parameters influences analysis accuracy.

The accuracy of three MRS analysis methods was compared: (1) ABfit, (2) ABfit-reg, and (3) LCModel, where ABfit-reg is a modified version of ABfit incorporating regularization. Accuracy was assessed on synthetic MRS data generated with random variability in the frequency shift and linewidth parameters applied to each basis signal. Spectra ( $$ ) were generated across a range of SNR values (10, 30, 60, 100) to evaluate the impact of variable data quality.

Comparison between ABfit and ABfit-reg demonstrates a statistically significant (p <  0.0005) improvement in accuracy associated with regularization for each SNR regime. An approximately 10% reduction in the mean squared metabolite errors was found for ABfit-reg compared to LCModel for SNR >10 (p <  0.0005). Furthermore, Bland-Altman analysis shows that incorporating regularization into ABfit enhances its agreement with LCModel.

Regularization is beneficial for MRS fitting and accurate characterization of the frequency and linewidth variability in vivo may yield further improvements.