Generative prior-guided neural interface reconstruction for 3D electrical impedance tomography
Journal of Computational Physics · 2026-07 · 557 · 114841
DOI: 10.1016/j.jcp.2026.114841 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Computational Physics, 2026-07. Verfasst von Haibo Liu und 2 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben. Gefördert von: Advanced Scientific Computing Research, National Key Research and Development Program of China, National Science Foundation Division of Electrical Communications and Cyber Systems.
Abstract (Originalsprache)
Reconstructing complex 3D interfaces from indirect measurements remains a grand challenge in scientific computing, particularly for ill-posed inverse problems like Electrical Impedance Tomography (EIT). Traditional shape optimization struggles with topological changes and regularization tuning, while emerging deep learning approaches often compromise physical fidelity or require prohibitive amounts of paired training data. We present a transformative ``solver-in-the-loop''framework that bridges this divide by coupling a pre-trained 3D generative prior with a rigorous boundary integral equation (BIE) solver. Unlike Physics-Informed Neural Networks (PINNs) that treat physics as soft constraints, our architecture enforces the governing elliptic PDE as a hard constraint at every optimization step, ensuring strict physical consistency. Simultaneously, we navigate a compact latent manifold of plausible geometries learned by a differentiable neural shape representation, effectively regularizing the ill-posed problem through data-driven priors rather than heuristic smoothing. By propagating adjoint shape derivatives directly through the neural decoder, we achieve fast, stable convergence with dramatically reduced degrees of freedom. Extensive experiments on 3D high-contrast EIT demonstrate that this principled hybrid approach yields superior geometric accuracy and data efficiency which is difficult to achieve using traditional methods, establishing a robust new paradigm for physics-constrained geometric discovery.
Förderung
- Advanced Scientific Computing Research
- National Key Research and Development Program of China
- National Science Foundation Division of Electrical Communications and Cyber Systems
- Computer and Information Science and Engineering Office of Advanced Cyberinfrastructure
- National Science Foundation Division of Mathematical Sciences
- National Science Foundation
- U.S. Department of Energy
- National Science Foundation Division of Chemical Bioengineering Environmental and Transport Systems
Zitieren
APA
Liu, H., Chen, J., & Lin, G. (2026). Generative prior-guided neural interface reconstruction for 3D electrical impedance tomography. Journal of Computational Physics, 557, 114841. https://doi.org/10.1016/j.jcp.2026.114841
Vancouver
Liu H, Chen J, Lin G. Generative prior-guided neural interface reconstruction for 3D electrical impedance tomography. Journal of Computational Physics. 2026;557:114841. doi:10.1016/j.jcp.2026.114841
BibTeX
@article{Liu2026,
author = {Liu, Haibo and Chen, Junqing and Lin, Guang},
title = {Generative prior-guided neural interface reconstruction for 3D electrical impedance tomography},
journal = {Journal of Computational Physics},
year = {2026},
volume = {557},
pages = {114841},
doi = {10.1016/j.jcp.2026.114841}
}
EUVIMED