Complex imaging of phase domains by deep neural networks

Longlong Wu; Pavol Juhas; Shinjae Yoo; Ian Robinson

doi:10.1107/S2052252520013780

Complex imaging of phase domains by deep neural networks

IUCrJ. 2021 Jan 1;8(Pt 1):12-21. doi: 10.1107/S2052252520013780.

Authors

Longlong Wu^{1

2}, Pavol Juhas¹, Shinjae Yoo¹, Ian Robinson^{2

3}

Affiliations

¹ Computational Science Initiative, Brookhaven National Laboratory, Upton, NY 11973, USA.
² Condensed Matter Physics and Materials Science Department, Brookhaven National Laboratory, Upton, NY 11973, USA.
³ London Centre for Nanotechnology, University College London, London, WC1E 6BT, United Kingdom.

Abstract

The reconstruction of a single-particle image from the modulus of its Fourier transform, by phase-retrieval methods, has been extensively applied in X-ray structural science. Particularly for strong-phase objects, such as the phase domains found inside crystals by Bragg coherent diffraction imaging (BCDI), conventional iteration methods are time consuming and sensitive to their initial guess because of their iterative nature. Here, a deep-neural-network model is presented which gives a fast and accurate estimate of the complex single-particle image in the form of a universal approximator learned from synthetic data. A way to combine the deep-neural-network model with conventional iterative methods is then presented to refine the accuracy of the reconstructed results from the proposed deep-neural-network model. Improved convergence is also demonstrated with experimental BCDI data.

Keywords: Bragg coherent X-ray diffraction; deep neural networks; machine learning; phase retrieval; single-particle imaging.

Grants and funding

This work was funded by U.S. Department of Energy, Office of Science grants DE-AC02-06CH11357 and DE-SC0012704. National Science Foundation grant DMR-9724294.