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 Experiment Video

Updated: May 6, 2026

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
08:20

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

2.9K

A Deep Learning Approach for Classifying Benign, Malignant, and Borderline Ovarian Tumors Using Convolutional Neural

Maria Giourga1, Ioannis Petropoulos2, Sofoklis Stavros3

  • 1First Department of Obstetrics and Gynecology, Alexandra Hospital, Medical School, National and Kapodistrian University of Athens, 11528 Athens, Greece.

Medical Sciences (Basel, Switzerland)
|February 20, 2026
PubMed
Summary

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

Endometrial Microbiome Profiles in Women Evaluated for Infertility or Recurrent Miscarriage: A Single-Center Descriptive Study.

Diagnostics (Basel, Switzerland)·2026
Same author

The Microbiome-Mitochondria-Extracellular Vesicle Axis in HPV Persistence and Cervical Carcinogenesis.

Genes·2026
Same author

Psychological Experiences of Pregnancy Following Transplantation: A Systematic Qualitative Review.

Medicina (Kaunas, Lithuania)·2026
Same author

Nutritional Challenges and Strategies in Obese Critically Ill Patients with Gynecological Cancer: A Narrative Review.

Nutrients·2026
Same author

Proteostasis is disrupted in human endothelial cells by serum from ATTR patients and is rescued by tafamidis treatment.

Amyloid : the international journal of experimental and clinical investigation : the official journal of the International Society of Amyloidosis·2026
Same author

Epigenetics, Oxidative Stress, and the Microbiome in Endometriosis: Toward an Integrated Mechanistic Framework for Precision Medicine.

Journal of personalized medicine·2026
This summary is machine-generated.

Generative Adversarial Networks (GANs) enhanced deep learning models can accurately classify ovarian masses, including borderline ovarian tumors (BOTs), using ultrasound images. This AI approach improves diagnostic accuracy and supports preoperative assessment.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Oncology

Background:

  • Accurate preoperative characterization of ovarian masses is crucial for patient management.
  • Borderline ovarian tumors (BOTs) present diagnostic challenges due to their rarity and ultrasound ambiguity.
  • Limited availability of expert sonologists necessitates advanced diagnostic tools.

Purpose of the Study:

  • To develop an AI classifier using Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs) for improved discrimination of ovarian masses.
  • To enhance the classification of benign, malignant, and borderline ovarian tumors (BOTs) from ultrasound images.
  • To address the class imbalance issue in deep learning models for rare diseases like BOTs.

Main Methods:

  • Retrospective analysis of 3816 ultrasound images from 636 ovarian masses (benign, malignant, BOTs).
Keywords:
artificial intelligencedeep convolutional generative adversarial network (DCGAN)deep learningovarian cancerovarian massultrasound imaging

Related Experiment Videos

Last Updated: May 6, 2026

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
08:20

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

2.9K
  • Utilized Deep Convolutional GAN (DCGAN) for data augmentation by generating 2000 synthetic BOT images.
  • Developed a three-class ensemble CNN model integrating VGG16, ResNet50, and InceptionNetV3 architectures.

    • Main Results:

    • DCGAN augmentation significantly improved BOT classification performance, increasing the F1-score from 68.4% to 86.5%.
    • Overall classification accuracy rose from 84.7% to 91.5% with DCGAN integration.
    • The final model achieved high sensitivity (88.2%) and specificity (85.1%) for BOT identification, with AUCs ranging from 0.91 to 0.96 across classes.

    Conclusions:

    • DCGAN-based data augmentation effectively overcomes dataset limitations for rare conditions like BOTs.
    • The enhanced CNN model shows significant potential as a decision support tool for preoperative ovarian mass assessment.
    • AI-driven approaches can improve diagnostic accuracy in challenging gynecological ultrasound cases.