Fast three-dimensional NMR spectroscopy reconstruction based on Hankel matrix with two-directional sliding windows

Area of Science:

• Biomedical analysis
• Spectroscopy
• Data reconstruction

Background:

• Nuclear Magnetic Resonance (NMR) spectroscopy is vital for biomedical analysis and quantification.
• High-dimensional NMR spectroscopy requires lengthy acquisition times, limiting its applications.
• Non-uniform sampling (NUS) reduces acquisition time but necessitates effective reconstruction methods.

Purpose of the Study:

• To develop a computationally efficient reconstruction method for high-dimensional NMR spectroscopy data acquired with NUS.
• To address the prolonged computational time associated with existing low-rank matrix reconstruction techniques.

Main Methods:

• Proposed a novel two-directional sliding window method to reduce the size of the low-rank matrix.
• Incorporated parallel computation to further accelerate the reconstruction process.
• Tested the method on synthetic and real NMR spectroscopy datasets.

Main Results:

• The proposed two-directional sliding window method significantly reduces computational time compared to traditional low-rank matrix reconstruction.
• Parallel computation further enhances the speed of the reconstruction process.
• The method achieves reconstruction in the shortest time among compared approaches without compromising spectral quality.

Conclusions:

• The novel two-directional sliding window method offers a computationally efficient solution for reconstructing NUS NMR data.
• This approach accelerates NMR data processing, enabling broader applications in biomedical analysis.
• The method provides high-quality spectral reconstruction with significantly reduced computational cost.