Revisiting aliasing noise to build more robust sparsity in nonuniform sampling 2D-NMR

Area of Science:

• Nuclear Magnetic Resonance (NMR) Spectroscopy
• Data Acquisition and Processing

Background:

• Nonuniform sampling (NUS) is crucial for accelerating 2D NMR experiments, especially for small molecules.
• Subsampling below 50% coverage in NUS can introduce artifacts, often referred to as sampling noise.
• Weak aliasing artifacts are a growing concern with increasingly sparse NUS schedules.

Purpose of the Study:

• To investigate the causes of artifacts in 1D NUS of 2D NMR spectra.
• To develop strategies for mitigating sampling noise and improving spectral quality in sparse NUS.
• To create practical 'one-click' NUS schedules for wider application.

Main Methods:

• Revisiting potential causes of artifacts in 1D NUS of 2D NMR.
• Implementing a convolutional filter-based screening approach to detect repeat sequences in sampling schedules.
• Evaluating the impact of initial uniform sampling periods on artifact reduction.
• Testing strategies on HSQC and LR-HSQMBC experiments.

Main Results:

• Repeat sequences in dense sampling regions of sparse NUS schedules cause aliasing artifacts.
• Sampling schedules with low proportions of repeat sequences exhibit significantly reduced artifacts.
• A short period (2%-4%) of initial uniform sampling effectively suppresses sampling noise.
• Robust spectral reconstructions using iterative soft thresholding (IST) are achieved even with sparse NUS.

Conclusions:

• Identifying and avoiding repeat sequences in NUS schedules is key to reducing spectral artifacts.
• Incorporating brief initial uniform sampling periods is an effective remediation strategy for sparse NUS.
• Developed 'one-click' schedules offer a practical solution for broader adoption of robust NUS techniques.