Towards single-shot coherent imaging via overlap-free ptychography
arXiv:2602.21361v1 Announce Type: cross Abstract: Ptychographic imaging at synchrotron and XFEL sources requires dense overlapping scans, limiting throughput and increasing dose. Extending coherent diffractive imaging to overlap-free operation on extended samples remains an open problem. Here, we extend PtychoPINN (O. Hoidn \emph{et al.}, \emph{Scientific Reports} \textbf{13}, 22789, 2023) to deliver \emph{overlap-free, single-shot} reconstructions in a Fresnel coherent diffraction imaging (CDI) geometry while also accelerating conventional multi-shot ptychography. The framework couples a differentiable forward model of coherent scattering with a Poisson photon-counting likelihood; real-space overlap enters as a tunable parameter via coordinate-based grouping rather than a hard requirement. On synthetic benchmarks, reconstructions remain accurate at low counts ($\sim\!10^4$ photons/frame), and overlap-free single-shot reconstruction with an experimental probe reaches amplitude structu
arXiv:2602.21361v1 Announce Type: cross Abstract: Ptychographic imaging at synchrotron and XFEL sources requires dense overlapping scans, limiting throughput and increasing dose. Extending coherent diffractive imaging to overlap-free operation on extended samples remains an open problem. Here, we extend PtychoPINN (O. Hoidn \emph{et al.}, \emph{Scientific Reports} \textbf{13}, 22789, 2023) to deliver \emph{overlap-free, single-shot} reconstructions in a Fresnel coherent diffraction imaging (CDI) geometry while also accelerating conventional multi-shot ptychography. The framework couples a differentiable forward model of coherent scattering with a Poisson photon-counting likelihood; real-space overlap enters as a tunable parameter via coordinate-based grouping rather than a hard requirement. On synthetic benchmarks, reconstructions remain accurate at low counts ($\sim\!10^4$ photons/frame), and overlap-free single-shot reconstruction with an experimental probe reaches amplitude structural similarity (SSIM) 0.904, compared with 0.968 for overlap-constrained reconstruction. Against a data-saturated supervised model with the same backbone (16,384 training images), PtychoPINN achieves higher SSIM with only 1,024 images and generalizes to unseen illumination profiles. Per-graphics processing unit (GPU) throughput is approximately $40\times$ that of least-squares maximum-likelihood (LSQ-ML) reconstruction at matched $128\times128$ resolution. These results, validated on experimental data from the Advanced Photon Source and the Linac Coherent Light Source, unify single-exposure Fresnel CDI and overlapped ptychography within one framework, supporting dose-efficient, high-throughput imaging at modern light sources.
Executive Summary
This article presents a novel framework, PtychoPINN, for achieving overlap-free, single-shot coherent imaging via ptychography. By leveraging a differentiable forward model and a Poisson photon-counting likelihood, the framework enables reconstructions on extended samples, accelerating conventional multi-shot ptychography. The authors demonstrate the effectiveness of PtychoPINN on synthetic benchmarks and experimental data from leading light sources, achieving higher throughput and similar quality to traditional methods. This breakthrough has significant implications for high-throughput imaging at modern light sources, enabling dose-efficient data acquisition. The framework's flexibility and scalability make it an attractive solution for various applications in materials science and biology.
Key Points
- ▸ PtychoPINN framework achieves overlap-free, single-shot coherent imaging via ptychography
- ▸ Differentiable forward model and Poisson photon-counting likelihood enable reconstructions on extended samples
- ▸ Accelerated conventional multi-shot ptychography and higher throughput compared to traditional methods
Merits
Strength in Framework Design
The authors' innovative combination of a differentiable forward model and Poisson photon-counting likelihood allows for efficient and accurate reconstructions, making PtychoPINN a robust framework for various applications.
Advancements in High-Throughput Imaging
PtychoPINN's ability to accelerate conventional multi-shot ptychography and achieve higher throughput compared to traditional methods has significant implications for high-throughput imaging at modern light sources.
Demerits
Limited Experimental Validation
The article primarily relies on synthetic benchmarks and experimental data from leading light sources, limiting the scope of experimental validation and potential applications.
Potential Computational Complexity
The framework's reliance on complex differentiable models and Poisson photon-counting likelihood may introduce computational complexity, potentially limiting its adoption in resource-constrained environments.
Expert Commentary
This article presents a significant advancement in the field of coherent imaging, leveraging machine learning and deep learning algorithms to achieve overlap-free, single-shot reconstructions. The authors' innovative framework, PtychoPINN, demonstrates its potential for high-throughput imaging at modern light sources, with significant implications for materials science and biology. While the article's primary focus on synthetic benchmarks and experimental data from leading light sources may limit its scope, the framework's flexibility and scalability make it an attractive solution for various applications. As the field continues to evolve, it is essential to consider the broader implications of this breakthrough, including its potential impact on policy, funding priorities, and the future of imaging science.
Recommendations
- ✓ Funding agencies and policymakers should prioritize investments in advanced light sources and infrastructure to support the development and application of PtychoPINN and similar frameworks.
- ✓ Researchers should explore the potential applications of PtychoPINN in various fields, including materials science, biology, and medicine, to fully realize its benefits.
