Evaluating metrics of spectral quality in nonuniform sampling

Area of Science:

• Nuclear Magnetic Resonance (NMR) Spectroscopy
• Data Acquisition Strategies
• Signal Processing

Background:

• Nonuniform sampling (NUS) accelerates data acquisition in NMR but requires careful strategy.
• Adaptive approaches to NUS aim to optimize data collection dynamically.
• Assessing spectral quality and determining data collection endpoints are key challenges.

Purpose of the Study:

• To develop an adaptive approach for nonuniform sampling (NUS) in NMR spectroscopy.
• To evaluate the effectiveness of in situ receiver operator characteristic (IROC) for assessing spectral quality and defining adaptive sampling stop criteria.
• To investigate whether peak-to-sidelobe ratio (PSR), an a priori metric, can predict spectral quality determined by IROC and guide adaptive sampling.

Main Methods:

• Augmented the Nonuniform Sampling Contest (NUScon) workflow with an in situ receiver operator characteristic (IROC) module for spectral quality assessment.
• Assessed the predictive ability of peak-to-sidelobe ratio (PSR) for spectral quality, using IROC as the ground truth.
• Analyzed the relationship between seed optimization for NUS schedules and resulting spectral quality evaluated by IROC.

Main Results:

• The integrated IROC module effectively defines stop criteria for adaptive FID (Free Induction Decay) collection in NUS experiments.
• Peak-to-sidelobe ratio (PSR) optimization of NUS sampling schedules did not lead to improved spectral quality as assessed by IROC.
• Observed trends in IROC-reported spectral quality provide insights for designing future adaptive FID selection strategies.

Conclusions:

• Adaptive NUS strategies can be enhanced by incorporating IROC for quantitative spectral quality assessment and defining adaptive stop criteria.
• A priori metrics like PSR are insufficient for predicting spectral quality in adaptive NUS, and seed optimization based on PSR is not beneficial.
• Future adaptive sampling designs should leverage IROC-based quality assessments for more effective FID selection.