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

You might also read

Related Articles

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

Sort by
Same author

Comprehensive Candidate Gene for Ginsenoside Rg1 Biosynthesis: Identification, Systematic Analysis, and Verification.

Plants (Basel, Switzerland)·2026
Same author

Integrated Metabolomics and Selection Signal Analysis Provide Insights Into the Selection for Flavonol Biosynthesis Associated With Lettuce Quality Improvement.

Plant biotechnology journal·2026
Same author

RNF2 mediates H2A ubiquitination to promote colitis via suppressing monocyte-macrophage transition in mice.

Nature communications·2026
Same author

An interpretable memory forensics framework for unknown attack identification in power grid edge devices.

Scientific reports·2026
Same author

Oil-Water Flow Monitoring in Wellbores with Inflow Control Valves Using Distributed Acoustic Sensing.

Sensors (Basel, Switzerland)·2026
Same author

Ultrasound Domain Adaptation for Robust Kidney Segmentation via Spectral-Similarity-Guided Translation.

Journal of imaging informatics in medicine·2026

Related Experiment Video

Updated: Aug 29, 2025

Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
08:05

Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia

Published on: December 19, 2020

14.3K

An efficient deep neural network framework for COVID-19 lung infection segmentation.

Ge Jin1, Chuancai Liu1,2, Xu Chen1

  • 1School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, China.

Information Sciences
|September 7, 2022
PubMed
Summary

This study introduces a new deep learning model for segmenting pneumonia in CT scans, reducing physician workload. The novel method improves accuracy by using Vector Quantized Variational AutoEncoder and adversarial learning for better feature representation and handling class imbalance.

Keywords:
Adversarial networkCOVID-19Infection segmentationProportions lossSemi-supervised learningVQ-VAE

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

2.9K
Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

9.2K

Related Experiment Videos

Last Updated: Aug 29, 2025

Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
08:05

Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia

Published on: December 19, 2020

14.3K
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

2.9K
Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

9.2K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computer Vision

Background:

  • Coronavirus Disease 2019 (COVID-19) significantly impacted global health systems since 2020.
  • Automated segmentation of pneumonia lesions in CT images using deep learning offers potential to aid physicians and enhance diagnostics.
  • Challenges remain in acquiring high-quality annotations and addressing subtle inter-class differences in medical image segmentation.

Purpose of the Study:

  • To propose a novel deep neural network for automatic segmentation of infected areas in CT images.
  • To reduce annotation costs and improve feature representation using a Vector Quantized Variational AutoEncoder (VQ-VAE).
  • To enhance model generalization and mitigate class imbalance with a novel proportions loss and adversarial learning.

Main Methods:

  • A Resnet-based deep neural network architecture was developed for CT image segmentation.
  • A VQ-VAE branch was integrated for image reconstruction, decoder regularization, and improved latent map feature representation.
  • A novel proportions loss function and a semi-supervised adversarial learning mechanism were implemented to address class imbalance and leverage unlabeled data.

Main Results:

  • The proposed deep learning model demonstrated superior performance in segmenting pneumonia lesions from CT images.
  • The integration of VQ-VAE and proportions loss effectively improved feature representation and handled class imbalance.
  • The semi-supervised adversarial learning component further regularized the network by utilizing information from unlabeled images.

Conclusions:

  • The novel deep neural network effectively segments pneumonia in CT images, addressing key challenges in annotation cost and class imbalance.
  • The proposed method shows significant potential to assist radiologists and improve the efficiency of diagnosing COVID-19 related pneumonia.
  • The study validates the superiority of the developed approach through extensive experiments on the COVID-SemiSeg dataset.