Classification of Quality Assessment of MR Spectroscopy Data: Comparing Quantitative and Qualitative Assessments

Motivation: Qualitative assessment of magnetic resonance spectroscopy (MRS) data is the standard for data quality assessment (DQA), however it's inefficient with high interobservational variability. Goal(s): The goal was to compare qualitative DQA with quantitative DQA (signal to noise ratio (SNR), linewidth (FWHM), and Cramer-Rao lower bounds (CRLB)), determining if quantitative measures alone are sufficient. Approach: 7,155 spectra were classified on a 5-measure scale with ratings 1 (acceptable) to 5 (rejected). Results: SNR was negatively correlated and FWHM and CRLB were positively correlated with qualitative DQA. Multiple cases showed spectra quantitatively acceptable but qualitatively rejected, demonstrating additional measures are needed for comprehensive DQA. Impact: Providing quantitative analysis for qualitative MR spectroscopy ratings when comparing SNR, FWHM, and CRLB variables will direct the process of quality analysis to demonstrate that additional quantitative DQA measures are needed to provide comprehensive DQA.