A Sparse Model-Inspired Deep Thresholding Network for Exponential Signal Reconstruction-Application in Fast Biological Spectroscopy

• Published in: IEEE transaction on neural networks and learning systems Vol. 34; no. 10; pp. 7578 - 7592
• Main Authors: Wang, Zi, Guo, Di, Tu, Zhangren, Huang, Yihui, Zhou, Yirong, Wang, Jian, Feng, Liubin, Lin, Donghai, You, Yongfu, Agback, Tatiana, Orekhov, Vladislav, Qu, Xiaobo
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.10.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: accelerated nmr-spectroscopy; Algorithms; Bioinformatics (Computational Biology); Bioinformatik (beräkningsbiologi); Biological system modeling; Cloud computing; Computer and Information Sciences; Computer architecture; Computer Science; Data models; Data- och informationsvetenskap (Datateknik); Deep learning; deep learning (DL); Engineering; exponential signal; fast sampling; Image reconstruction; Iterative methods; Machine learning; Medical Imaging; Medicinsk bildvetenskap; Neural networks; Nuclear magnetic resonance; Optimization; resolution; Signal processing; Signal reconstruction; Spectroscopy; Spectrum analysis; Synthetic data; Time-domain analysis
• ISSN: 2162-237X; 2162-2388; 2162-2388
• DOI: 10.1109/TNNLS.2022.3144580

The nonuniform sampling (NUS) is a powerful approach to enable fast acquisition but requires sophisticated reconstruction algorithms. Faithful reconstruction from partially sampled exponentials is highly expected in general signal processing and many applications. Deep learning (DL) has shown astonishing potential in this field, but many existing problems, such as lack of robustness and explainability, greatly limit its applications. In this work, by combining the merits of the sparse model-based optimization method and data-driven DL, we propose a DL architecture for spectra reconstruction from undersampled data, called MoDern. It follows the iterative reconstruction in solving a sparse model to build the neural network, and we elaborately design a learnable soft-thresholding to adaptively eliminate the spectrum artifacts introduced by undersampling. Extensive results on both synthetic and biological data show that MoDern enables more robust, high-fidelity, and ultrafast reconstruction than the state-of-the-art methods. Remarkably, MoDern has a small number of network parameters and is trained on solely synthetic data while generalizing well to biological data in various scenarios. Furthermore, we extend it to an open-access and easy-to-use cloud computing platform (XCloud-MoDern), contributing a promising strategy for further development of biological applications.