# A Physics-Inspired Deep Learning Framework with Polar Coordinate Attention for Ptychographic Imaging Source: [A Physics-Inspired Deep Learning Framework with Polar Coordinate Attention for Ptychographic Imaging]() Collection index: [llms.txt]() Data updated: 2025-10-25T07:39:47.123Z Authors: Han Yue, Jun Cheng, Yu-Xuan Ren, Chien-Chun Chen, Grant A. van Riessen, Philip Heng Wai Leong, Steve Feng Shu Published: 2024-11-25T06:49:59.000Z ## Abstract Ptychographic imaging confronts inherent challenges in applying deep learning for phase retrieval from diffraction patterns. Conventional neural architectures, both convolutional neural networks and Transformer-based methods, are optimized for natural images with Euclidean spatial neighborhood-based inductive biases that exhibit geometric mismatch with the concentric coherent patterns characteristic of diffraction data in reciprocal space. In this paper, we present PPN, a physics-inspired deep learning network with Polar Coordinate Attention (PoCA) for ptychographic imaging, that aligns neural inductive biases with diffraction physics through a dual-branch architecture separating local feature extraction from non-local coherence modeling. It consists of a PoCA mechanism that replaces Euclidean spatial priors with physically consistent radial-angular correlations. PPN outperforms existing end-to-end models, with spectral and spatial analysis confirming its greater preservation of high-frequency details. Notably, PPN maintains robust performance compared to iterative methods even at low overlap ratios, making it well suited for high-throughput imaging in real-world acquisition scenarios for samples with consistent structural characteristics. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Amplitude Reconstruction on Synthetic LTEM Noise-Free (test)](): Amplitude Reconstruction; Synthetic LTEM Noise-Free per-sample evaluation (test); MSE: 485. - [Amplitude Reconstruction on Synthetic LTEM Noisy per-sample evaluation (test)](): Amplitude Reconstruction; Synthetic LTEM Noisy per-sample evaluation (test); MSE: 5.09. - [Phase Reconstruction on Real-world experimental X-ray dataset (stitched full-field reconstructions)](): Phase Reconstruction; Real-world experimental X-ray dataset stitched full-field reconstructions; MSE: 326. - [Phase Reconstruction on Synthetic LTEM Noise-Free per-sample evaluation (test)](): Phase Reconstruction; Synthetic LTEM Noise-Free per-sample evaluation (test); MSE: 119. - [Phase Reconstruction on Synthetic LTEM Noisy per-sample evaluation (test)](): Phase Reconstruction; Synthetic LTEM Noisy per-sample evaluation (test); MSE (x10^-2): 1.21. - [Stitched-image Amplitude Reconstruction on Stitched-image Synthetic (noise-free)](): Stitched-image Amplitude Reconstruction; Stitched-image Synthetic (noise-free); MSE: 0.0236. - [Stitched-image Amplitude Reconstruction on Stitched-image Synthetic (noisy)](): Stitched-image Amplitude Reconstruction; Stitched-image Synthetic (noisy); MSE: 0.0235. - [Stitched-image Phase Reconstruction on Stitched-image Synthetic (Noise-Free)](): Stitched-image Phase Reconstruction; Stitched-image Synthetic (noise-free); MSE: 0.0073. - [Stitched-image Phase Reconstruction on Stitched-image Synthetic (Noisy)](): Stitched-image Phase Reconstruction; Stitched-image Synthetic (noisy); MSE: 0.72. - [Amplitude Reconstruction on Real-world experimental X-ray dataset](): Amplitude Reconstruction; Real-world experimental X-ray dataset stitched full-field reconstructions; MSE: 0.16. ## Original sources and implementation [Original paper]() [PDF]()