Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

4D-ONIX for reconstructing 3D movies from sparse X-ray projections via deep learning

Zhang, Yuhe LU ; Yao, Zisheng LU orcid ; Klöfkorn, Robert LU orcid ; Ritschel, Tobias and Villanueva-Perez, Pablo LU orcid (2025) In Communications Engineering 4(1).
Abstract

The X-ray flux from X-ray free-electron lasers and storage rings enables new spatiotemporal opportunities for studying in-situ and operando dynamics, even with single pulses. X-ray multi-projection imaging is a technique that provides volumetric information using single pulses while avoiding the centrifugal forces induced by conventional time-resolved 3D methods like time-resolved tomography, and can acquire 3D movies (4D) at least three orders of magnitude faster than existing techniques. However, reconstructing 4D information from highly sparse projections remains a challenge for current algorithms. Here we present 4D-ONIX, a deep-learning-based approach that reconstructs 3D movies from an extremely limited number of projections. It... (More)

The X-ray flux from X-ray free-electron lasers and storage rings enables new spatiotemporal opportunities for studying in-situ and operando dynamics, even with single pulses. X-ray multi-projection imaging is a technique that provides volumetric information using single pulses while avoiding the centrifugal forces induced by conventional time-resolved 3D methods like time-resolved tomography, and can acquire 3D movies (4D) at least three orders of magnitude faster than existing techniques. However, reconstructing 4D information from highly sparse projections remains a challenge for current algorithms. Here we present 4D-ONIX, a deep-learning-based approach that reconstructs 3D movies from an extremely limited number of projections. It combines the computational physical model of X-ray interaction with matter and state-of-the-art deep learning methods. We demonstrate its ability to reconstruct high-quality 4D by generalizing over multiple experiments with only two to three projections per timestamp on simulations of water droplet collisions and experimental data of additive manufacturing. Our results demonstrate 4D-ONIX as an enabling tool for 4D analysis, offering high-quality image reconstruction for fast dynamics three orders of magnitude faster than tomography.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
Communications Engineering
volume
4
issue
1
article number
54
publisher
Nature Publishing Group UK London
external identifiers
  • scopus:105000506690
  • pmid:40119014
ISSN
2731-3395
DOI
10.1038/s44172-025-00390-w
language
English
LU publication?
yes
id
ab751905-713d-4a61-9ef2-1ee11b5b3bf3
date added to LUP
2025-08-07 10:34:21
date last changed
2025-08-08 03:30:25
@article{ab751905-713d-4a61-9ef2-1ee11b5b3bf3,
  abstract     = {{<p>The X-ray flux from X-ray free-electron lasers and storage rings enables new spatiotemporal opportunities for studying in-situ and operando dynamics, even with single pulses. X-ray multi-projection imaging is a technique that provides volumetric information using single pulses while avoiding the centrifugal forces induced by conventional time-resolved 3D methods like time-resolved tomography, and can acquire 3D movies (4D) at least three orders of magnitude faster than existing techniques. However, reconstructing 4D information from highly sparse projections remains a challenge for current algorithms. Here we present 4D-ONIX, a deep-learning-based approach that reconstructs 3D movies from an extremely limited number of projections. It combines the computational physical model of X-ray interaction with matter and state-of-the-art deep learning methods. We demonstrate its ability to reconstruct high-quality 4D by generalizing over multiple experiments with only two to three projections per timestamp on simulations of water droplet collisions and experimental data of additive manufacturing. Our results demonstrate 4D-ONIX as an enabling tool for 4D analysis, offering high-quality image reconstruction for fast dynamics three orders of magnitude faster than tomography.</p>}},
  author       = {{Zhang, Yuhe and Yao, Zisheng and Klöfkorn, Robert and Ritschel, Tobias and Villanueva-Perez, Pablo}},
  issn         = {{2731-3395}},
  language     = {{eng}},
  number       = {{1}},
  publisher    = {{Nature Publishing Group UK London}},
  series       = {{Communications Engineering}},
  title        = {{4D-ONIX for reconstructing 3D movies from sparse X-ray projections via deep learning}},
  url          = {{http://dx.doi.org/10.1038/s44172-025-00390-w}},
  doi          = {{10.1038/s44172-025-00390-w}},
  volume       = {{4}},
  year         = {{2025}},
}