Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Super-resolution Fluorescence Microscopy01:37

Super-resolution Fluorescence Microscopy

7.0K
Super-resolution fluorescence microscopy (SRFM) provides a better resolution than conventional fluorescence microscopy by reducing the point spread function (PSF). PSF is the light intensity distribution from a point that causes it to appear blurred. Due to PSF, each fluorescing point appears bigger than its actual size, and it is the PSF interference of nearby fluorophores that causes the blurred image. Various approaches to achieving higher resolution through SRFM have recently been...
7.0K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Generative AI for misalignment-resistant virtual staining to accelerate histopathology workflows.

Nature communications·2026
Same author

<i>Lepidium virginicum</i> Water-Soluble Chlorophyll-Binding Protein with Chlorophyll A as a Novel Contrast Agent for Photoacoustic Imaging.

Sensors (Basel, Switzerland)·2025
Same author

Deep learning-assisted low-cost autofluorescence microscopy for rapid slide-free imaging with virtual histological staining.

Biomedical optics express·2024
Same author

Deep learning-assisted smartphone-based quantitative microscopy for label-free peripheral blood smear analysis.

Biomedical optics express·2024
Same author

Review of low-cost light sources and miniaturized designs in photoacoustic microscopy.

Journal of biomedical optics·2023
Same author

Translational rapid ultraviolet-excited sectioning tomography for whole-organ multicolor imaging with real-time molecular staining.

eLife·2022

Related Experiment Video

Updated: Jun 28, 2025

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

394

Exceeding the limit for microscopic image translation with a deep learning-based unified framework.

Weixing Dai1, Ivy H M Wong1, Terence T W Wong1

  • 1Department of Chemical and Biological Engineering, Translational and Advanced Bioimaging Laboratory, Hong Kong University of Science and Technology, Hong Kong 999077, China.

PNAS Nexus
|April 11, 2024
PubMed
Summary
This summary is machine-generated.

A new unified framework (U-Frame) handles microscopic image translation with roughly paired data. This approach outperforms existing supervised and unsupervised methods across various applications, even with perfectly aligned images.

Keywords:
deep learningmicroscopic image translationsupervised learningunified frameworkunsupervised learning

More Related Videos

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.7K
Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

8.8K

Related Experiment Videos

Last Updated: Jun 28, 2025

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

394
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.7K
Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

8.8K

Area of Science:

  • Microscopy
  • Artificial Intelligence
  • Image Processing

Background:

  • Deep learning models for microscopic image translation typically require either perfectly paired data for supervised learning or ignore pairing information in unsupervised learning.
  • Real-world datasets often exhibit rough data pairing, limiting the effectiveness of standard supervised and unsupervised deep learning approaches.
  • Existing methods struggle when data alignment is imperfect, leading to suboptimal performance in microscopic image translation tasks.

Purpose of the Study:

  • To propose a unified framework (U-Frame) that integrates supervised and unsupervised learning for microscopic image translation.
  • To address the challenge of roughly paired data by introducing an adaptive tolerance size and global sampling rule.
  • To demonstrate the superiority of U-Frame over conventional methods across diverse image translation applications and varying degrees of data misalignment.

Main Methods:

  • Developed a unified framework (U-Frame) that combines supervised and unsupervised learning principles.
  • Implemented an automatically adjustable tolerance size to accommodate varying degrees of data misalignment.
  • Incorporated a global sampling rule to effectively utilize roughly paired image data.

Main Results:

  • U-Frame consistently outperformed both purely supervised and unsupervised learning methods across all tested levels of data misalignment.
  • The framework demonstrated robust performance even on perfectly aligned image pairs, indicating its versatility.
  • Achieved significant improvements in various applications, including pseudo-optical sectioning, virtual histological staining, image quality enhancement, and fluorescent label prediction.

Conclusions:

  • The proposed U-Frame offers a robust and versatile solution for microscopic image translation, effectively handling imperfectly paired data.
  • U-Frame sets a new standard for image translation tasks by unifying learning paradigms and improving performance across diverse applications.
  • The framework's ability to adapt to misalignment and leverage available pairing information makes it a valuable tool for scientific image analysis and diagnostics.