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 6, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

Multiple-instance learning algorithms for computer-aided detection.

M Murat Dundar1, Glenn Fung, Balaji Krishnapuram

  • 1Computer Aided Diagnosis and Knowledge Solutions, Siemens Medical Solutions, 51 Valley Stream Parkway, MS E51, Malvern, PA 19355, USA. murat.dundar@siemens.com

IEEE Transactions on Bio-Medical Engineering
|March 13, 2008
PubMed
Summary

A new Convex Hull (CH) framework significantly speeds up multiple-instance learning (MIL) for computer-aided diagnosis (CAD) with unbalanced data. This efficient approach improves diagnostic accuracy compared to existing methods.

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

Identification of predictive patient characteristics for assessing the probability of COVID-19 in-hospital mortality.

PLOS digital health·2024
Same author

Deep-Learning-Based High-Intensity Focused Ultrasound Lesion Segmentation in Multi-Wavelength Photoacoustic Imaging.

Bioengineering (Basel, Switzerland)·2023
Same author

A machine learning model for orthodontic extraction/non-extraction decision in a racially and ethnically diverse patient population.

International orthodontics·2023
Same author

Multi Resolution Analysis (MRA) for Approximate Self-Attention.

Proceedings of machine learning research·2023
Same author

A novel statistical methodology for quantifying the spatial arrangements of axons in peripheral nerves.

Frontiers in neuroscience·2023
Same author

Can we predict orthodontic extraction patterns by using machine learning?

Orthodontics & craniofacial research·2023

Area of Science:

  • Computer Science
  • Machine Learning
  • Medical Imaging

Background:

  • Computer-aided diagnosis (CAD) often involves multiple-instance learning (MIL) with highly unbalanced datasets.
  • Existing MIL algorithms are computationally expensive for these large, imbalanced datasets.

Purpose of the Study:

  • To introduce a computationally efficient framework for MIL problems with unbalanced data.
  • To improve diagnostic accuracy in computer-aided diagnosis applications.

Main Methods:

  • Developed the Convex Hull (CH) framework for learning multiple-instance representations.
  • The CH framework is compatible with standard hyperplane-based learning algorithms.
  • Guarantees global optimal solutions for certain algorithms.

More Related Videos

Superior Auto-Identification of Trypanosome Parasites by Using a Hybrid Deep-Learning Model
08:20

Superior Auto-Identification of Trypanosome Parasites by Using a Hybrid Deep-Learning Model

Published on: October 27, 2023

Related Experiment Videos

Last Updated: Jul 6, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
03:31

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

Superior Auto-Identification of Trypanosome Parasites by Using a Hybrid Deep-Learning Model
08:20

Superior Auto-Identification of Trypanosome Parasites by Using a Hybrid Deep-Learning Model

Published on: October 27, 2023

Main Results:

  • The CH framework demonstrates significantly faster computation compared to existing MIL algorithms.
  • Experimental studies on two CAD applications show improved diagnostic accuracy.
  • Outperforms both traditional classifiers and other MIL methods on benchmark datasets.

Conclusions:

  • The CH framework offers a computationally efficient and accurate solution for MIL problems in CAD.
  • It provides a competitive alternative to state-of-the-art MIL methods, especially for unbalanced data.