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: Jul 7, 2026

Synchronous Triplanar Reconstruction Integrated with Color Doppler Mapping for Precise and Rapid Localization of Thyroid Lesions
05:41

Synchronous Triplanar Reconstruction Integrated with Color Doppler Mapping for Precise and Rapid Localization of Thyroid Lesions

Published on: February 9, 2024

Acquiring Weak Annotations for Tumor Localization in Temporal and Volumetric Data.

Yu-Cheng Chou1, Bowen Li1, Deng-Ping Fan2

  • 1Department of Computer Science, Johns Hopkins University, Baltimore 21218, USA.

Machine Intelligence Research (Beijing)
|July 6, 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

Hybrid Curriculum Learning for Data-Efficient Lung Nodule Detection with YOLOv11.

Diagnostics (Basel, Switzerland)·2026
Same author

Hyperplasia Functions as a Link between Obesity and Cancer.

Cancer research·2026
See all related articles

A new Drag&Drop annotation strategy efficiently creates large datasets for AI-powered tumor detection. This method rivals detailed annotations, improving AI model robustness with diverse, weakly labeled data.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computer Vision

Background:

  • Automated tumor detection and localization rely on large, well-annotated datasets for AI training.
  • Pixel-wise annotation is time-consuming and resource-intensive, especially for high-dimensional medical data like colonoscopy videos and CT scans.

Purpose of the Study:

  • To introduce an efficient annotation strategy, Drag&Drop, for medical imaging AI.
  • To develop a weakly supervised learning method compatible with Drag&Drop annotations.
  • To evaluate the performance and robustness of the proposed annotation strategy and learning method.

Main Methods:

  • Developed the Drag&Drop annotation strategy, simplifying annotation to a drag-and-drop process.
  • Proposed a novel weakly supervised learning method utilizing the watershed algorithm for Drag&Drop annotations.
Keywords:
Weak annotationabdomencolonoscopydetectionlocalizationsegmentation

More Related Videos

Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment
13:01

Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment

Published on: June 3, 2022

Related Experiment Videos

Last Updated: Jul 7, 2026

Synchronous Triplanar Reconstruction Integrated with Color Doppler Mapping for Precise and Rapid Localization of Thyroid Lesions
05:41

Synchronous Triplanar Reconstruction Integrated with Color Doppler Mapping for Precise and Rapid Localization of Thyroid Lesions

Published on: February 9, 2024

Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment
13:01

Industrialized, Artificial Intelligence-guided Laser Microdissection for Microscaled Proteomic Analysis of the Tumor Microenvironment

Published on: June 3, 2022

  • Conducted experiments on colonoscopy videos (polyps) and CT scans (pancreatic tumors).

    • Main Results:

    • The Drag&Drop strategy proved more efficient than traditional weak annotations (bounding boxes, scribbles) for temporal and volumetric data.
    • The developed weakly supervised method achieved comparable performance to models trained on detailed per-pixel annotations.
    • Weak annotations from diverse patient data enhanced model robustness to unseen images more effectively than limited per-pixel annotations.

    Conclusions:

    • The Drag&Drop strategy offers an efficient approach for creating large-scale medical datasets for tumor screening.
    • This method provides a practical solution for resource-limited scenarios, balancing annotation effort with model performance.
    • Weakly supervised learning with diverse, efficiently annotated data can yield robust AI models for medical image analysis across modalities.