A Physics-Inspired Deep Learning Framework with Polar Coordinate Attention for Ptychographic Imaging

2024-11-25

AI Summary

In ptychographic imaging, conventional deep learning methods—such as CNNs and Transformers—are hindered by geometric mismatch between their Euclidean-space inductive biases and the concentric reciprocal-space structure of diffraction patterns, limiting phase retrieval performance. To address this, we propose the physics-driven Ptychographic Prior Network (PPN), featuring a novel Polar Coordinate Attention (PoCA) mechanism that explicitly models radial–angular coherence in diffraction data. PPN employs a dual-branch encoder–decoder architecture to decouple local feature extraction from non-local coherence modeling, ensuring consistent alignment between neural priors and physical constraints. The network demonstrates robustness under low probe overlap and achieves significantly higher spectral fidelity than state-of-the-art end-to-end models, enabling high-throughput, real-world ptychographic imaging.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search
• Machine Learning: Deep Neural Architectures and Foundation Models

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Security and Privacy: Data transparency and provenance

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.

Problem

• Addresses geometric mismatch in deep learning for ptychographic phase retrieval
• Replaces Euclidean spatial priors with polar coordinate attention for diffraction physics
• Enhances high-frequency detail preservation and robustness at low overlap ratios

Innovation

• Polar Coordinate Attention for diffraction physics alignment
• Dual-branch architecture separates local and non-local features
• Radial-angular correlations replace Euclidean spatial priors