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

  1. Home
  3. Research Domains
  5. Information And Computing Sciences
  7. Artificial Intelligence
  9. Knowledge Representation And Reasoning
  11. Semantic-driven Synthesis Of Histological Images With Controllable Cellular Distributions.

Semantic-driven synthesis of histological images with controllable cellular distributions.

Alen Shahini1, Alessandro Gambella2, Filippo Molinari1

  • 1Biolab, PoliTo(BIO)Med Lab, Department of Electronics and Telecommunications, Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 Turin, Italy.

Computer Methods and Programs in Biomedicine
|January 31, 2025

Related Experiment Videos

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone
09:31

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone

Published on: April 8, 2015

11.5K
Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research
05:22

Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research

Published on: June 21, 2024

309
Multimodal Hierarchical Imaging of Serial Sections for Finding Specific Cellular Targets within Large Volumes
11:19

Multimodal Hierarchical Imaging of Serial Sections for Finding Specific Cellular Targets within Large Volumes

Published on: March 20, 2018

10.4K

View abstract on PubMed

Summary
This summary is machine-generated.

We developed SENSE, a framework to create realistic synthetic histology images for digital pathology training. This tool enhances computational pathology datasets, improving cell detection and analysis accuracy.

Area of Science:

  • Computational pathology
  • Digital pathology
  • Medical image synthesis

Background:

  • Digital pathology requires large, annotated datasets for computational method training.
  • Dataset generation is hindered by expert annotation needs and operator variability.

Purpose of the Study:

  • To introduce SENSE (SEmantic Nuclear Synthesis Emulator), a novel framework for synthesizing realistic histological images.
  • To enable precise control over cellular distributions and properties in synthetic images.
  • To address challenges in creating high-quality datasets for computational pathology.

Main Methods:

  • Developed a statistical modeling system for class-specific nuclear characteristics.
  • Employed a hybrid Vision Transformer (ViT)-Pix2Pix Generative Adversarial Network (GAN) architecture.
Keywords:
Digital pathologyGenerative adversarial networksImage simulationInstance-aware models

Related Experiment Videos

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone
09:31

Automated Quantification of Hematopoietic Cell – Stromal Cell Interactions in Histological Images of Undecalcified Bone

Published on: April 8, 2015

11.5K
Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research
05:22

Author Spotlight: Enhanced Multiplex Immunofluorescent Microscopy Protocol for Neuroscience Research

Published on: June 21, 2024

309
Multimodal Hierarchical Imaging of Serial Sections for Finding Specific Cellular Targets within Large Volumes
11:19

Multimodal Hierarchical Imaging of Serial Sections for Finding Specific Cellular Targets within Large Volumes

Published on: March 20, 2018

10.4K
  • Implemented a modular design for independent control of cellular type, count, and spatial distribution.

    • Main Results:

    • SENSE-generated images matched real sample quality (MANIQA: 0.52 ± 0.03 vs 0.52 ± 0.04).
    • Biological plausibility was maintained, verified by experts.
    • Augmenting training data with synthetic images improved segmentation performance (DSC from 79.71 to 84.86).
    • Neutrophil segmentation accuracy increased significantly (40.18 to 78.71 DSC).

    Conclusions:

    • SENSE enables targeted dataset enhancement for computational pathology.
    • The framework offers new possibilities for educational and training scenarios.
    • SENSE facilitates the creation of diverse, biologically plausible histological image datasets.
    Multi-scale models