Abstract:Fourier ptychographic microscopy (FPM) is a promising imaging technique which can circumvent the space‐bandwidth product (SBP) of the system and achieve a reconstruction result with wide field‐of‐view (FOV), high‐resolution and quantitative phase information. However, traditional iterative‐based methods typically require multiple times to get convergence, and due to the wave vector deviation in different areas, the millimeter‐level full‐FOV cannot be well reconstructed once and typically required to be separated into several portions with sufficient overlaps and reconstructed separately, which makes traditional methods suffer from long reconstruction time for a large‐FOV (of the order of minutes) and limits the application in real‐time large‐FOV monitoring of live sample in vitro. Here we propose a novel deep‐learning based method called DFNN which can be used in place of traditional iterative‐based methods to increase the quality of single large‐FOV reconstruction and reducing the processing time from 167.5 to 0.1125 second. In addition, we demonstrate that by training based on the simulation dataset with 'high‐entropy property [Opt. Express 28, 24152‐241720 (2020)], DFNN could has fine generalizability and little dependence on the morphological features of samples. The superior robustness of DFNN against noise is also demonstrated in both simulation and experiment. Furthermore, our model shows more robustness against the wave vector deviation. therefore, we could achieve better results at the edge areas of a single large‐FOV reconstruction. Our method demonstrates a promising way to perform real‐time single large‐FOV reconstructions and provides further possibilities for real‐time large‐FOV monitoring of live samples with sub‐cellular resolution.This article is protected by copyright. All rights reserved.

Wide-field, high-resolution reconstruction in computational multi-aperture miniscope using a Fourier neural network

Wide-Field, High-Resolution Reconstruction in Computational Multi-Aperture Miniscope Using a Fourier Neural Network

Chip-based Wide-Field 3D Nanoscopy Through Tunable Spatial-Frequency-shift Effect

Untrained Neural Network Enabling Fast and Universal Structured-Illumination Microscopy

Wide-field 3D Nanoscopy on Chip Through Large and Tunable Spatial-Frequency-shift Effect

Three-dimensional Multi-Color Optical Nanoscopy at Sub-10-nm Resolution Based on Small-Molecule Organic Probes.

Universal and High-Fidelity Resolution Extending for Fluorescence Microscopy Using a Single-Training Physics-Informed Sparse Neural Network

Extended Depth of Field Reconstruction with Complex Field Estimation

Advancing Computational Miniature Mesoscope with Simulator-based Deep Learning Reconstruction

Double‐flow convolutional neural network for rapid large field of view Fourier ptychographic reconstruction

A Physics-Inspired Deep Learning Framework for an Efficient Fourier Ptychographic Microscopy Reconstruction under Low Overlap Conditions

Single-Shot 3D Widefield Fluorescence Imaging with a Computational Miniature Mesoscope

High-Throughput Deep Learning Microscopy Using Multi-Angle Super-Resolution

Deep-learning Light-Field Microscopy with Improved Resolution and Reconstruction Speed

A miniature fluorescence microscope for multi-plane imaging

Enhance imaging depth in wide-field two-photon microscopy by extended detection and computational reconstruction

Multifocal fluorescence video-rate imaging of centimetre-wide arbitrarily shaped brain surfaces at micrometric resolution

Robust single-shot 3D fluorescence imaging in scattering media with a simulator-trained neural network

Rapid and Extreme Low-light Superresolution Imaging Via Artificial Intelligence

Real-Time Resolution Enhancement of Confocal Laser Scanning Microscopy via Deep Learning

Three-dimensional reconstruction based on micro-imaging under wavelength-tunable illumination