Physics-aware resolution enhancement of soft X-ray tomography with measurement-supervised learning
- 1University College Dublin
Abstract
Soft X-ray tomography (SXT) is an emerging modality for whole-cell 3D imaging in near-native states. However, the effective spatial resolution is limited by optical artifacts characterized by the point spread function (PSF). Standard reconstruction methods force a compromise between structural sharpness and noise, failing to fully resolve these depth-dependent artifacts. By embedding experimentally measured, depth-variant PSFs into a differentiable forward model, we demonstrate a physics-aware computational optimization that bypasses these limitations to recover high-frequency cellular ultrastructure. The structural fidelity was validated using split-tilt Fourier ring correlation (FRC), alongside an experimental bead phantom tomogram, providing supporting evidence that the recovered high-frequency features reflect genuine specimen structure rather than fabricated artifacts. Our method effectively increases FRC spatial resolution and recovers cellular ultrastructure. Furthermore, under sparse-angular subsampling, the framework maintained spatial resolution using half the projection angles, a computational proxy pointing toward the potential for reduced radiation exposure in future acquisitions. This hardware-free, computational approach offers a route toward mitigating the optical and dosimetric constraints that currently limit nanoscale soft X-ray tomography.
Related
- HistoTx: Early fusion of H&E images and spatial transcriptomics at varying spatial transcriptomics resolution with self-supervised learning
- Semi-supervised clustering with knowledge-guided representation learning in cryo-electron tomography
- Excluded Volume Effect-induced Shrinkage for Resolution Enhancement of Hydrogel Bioconstructs
- Ultrahigh-resolution ophthalmic optical coherence tomography
- Whole-body Super-resolution Functional and Molecular Imaging with Panoramic Photoacoustic-Ultrasound Tomography
BibTeX
@misc{chueh2026physicsaware,
title = {{Physics-aware resolution enhancement of soft X-ray tomography with measurement-supervised learning}},
author = {S. Chueh and E. Gallagher and C. de Ceuninck van Capelle and L. Luo and T. Ishikawa and C. Evans and N. Fletcher and M. Lopez-Perez and D. Rogers and S. O'Connor and C. McIntyre and M. Donnellan and P. Sheridan and J. C. Simpson and S. Kapishnikov},
howpublished = {bioRxiv},
year = {2026},
month = jun,
doi = {10.64898/2026.06.21.730079},
url = {https://www.biorxiv.org/content/10.64898/2026.06.21.730079v3},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Physics-aware resolution enhancement of soft X-ray tomography with measurement-supervised learning
AU - S. Chueh
AU - E. Gallagher
AU - C. de Ceuninck van Capelle
AU - L. Luo
AU - T. Ishikawa
AU - C. Evans
AU - N. Fletcher
AU - M. Lopez-Perez
AU - D. Rogers
AU - S. O'Connor
AU - C. McIntyre
AU - M. Donnellan
AU - P. Sheridan
AU - J. C. Simpson
AU - S. Kapishnikov
PY - 2026
DA - 2026/06/21
PB - bioRxiv
DO - 10.64898/2026.06.21.730079
UR - https://www.biorxiv.org/content/10.64898/2026.06.21.730079v3
AB - Soft X-ray tomography (SXT) is an emerging modality for whole-cell 3D imaging in near-native states. However, the effective spatial resolution is limited by optical artifacts characterized by the point spread function (PSF). Standard reconstruction methods force a compromise between structural sharpness and noise, failing to fully resolve these depth-dependent artifacts. By embedding experimentally measured, depth-variant PSFs into a differentiable forward model, we demonstrate a physics-aware computational optimization that bypasses these limitations to recover high-frequency cellular ultrastructure. The structural fidelity was validated using split-tilt Fourier ring correlation (FRC), alongside an experimental bead phantom tomogram, providing supporting evidence that the recovered high-frequency features reflect genuine specimen structure rather than fabricated artifacts. Our method effectively increases FRC spatial resolution and recovers cellular ultrastructure. Furthermore, under sparse-angular subsampling, the framework maintained spatial resolution using half the projection angles, a computational proxy pointing toward the potential for reduced radiation exposure in future acquisitions. This hardware-free, computational approach offers a route toward mitigating the optical and dosimetric constraints that currently limit nanoscale soft X-ray tomography.
ER -