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 1, 2026

Design and Analysis for Fall Detection System Simplification
08:05

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

Strategies for Class-Imbalanced Learning in Multi-Sensor Medical Imaging.

Da Zhou1,2,3, Song Gao2,4, Xinrui Huang1,2,3

  • 1Department of Biophysics, School of Basic Medical Sciences, Peking University, Beijing 100191, China.

Sensors (Basel, Switzerland)
|March 28, 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

γ-GC mitigates MASLD by reducing CD36 palmitoylation and plasma membrane translocation mediated by AKT/KLF10/zDHHC7 axis.

Nutrition & metabolism·2026
Same author

Intravenous human umbilical cord-derived mesenchymal stem cell transplantation for ischemic stroke (MERIT): A phase 1 trial.

Cell reports. Medicine·2026
Same author

An Association Among Problematic Smartphone Use, Sleep Quality, Behavioral Engagement and Academic Achievements: A Multiple Mediation Model Based on Large-Scale Data.

Behavioral sciences (Basel, Switzerland)·2026
Same author

Multi-kingdom profiling reveals altered gut phage-bacteria-metabolite interactions in MASLD.

Nature communications·2026
Same author

Efficacy of pancreatic enzyme replacement therapy (PERT) in the treatment of gastrointestinal symptoms after gastric cancer surgery: study protocol for a multicenter randomized controlled open-label trial.

Frontiers in pharmacology·2026
Same author

B4GALT1 deficiency attenuates steatohepatitis by regulating the PPARγ/ACSL4 axis.

Hepatology communications·2026

Addressing class imbalance in AI medical imaging is crucial. This review details data-centric and model-centric strategies, including multi-sensor fusion, to improve diagnostic accuracy for rare conditions.

Area of Science:

  • Medical Imaging AI
  • Machine Learning in Healthcare

Background:

  • Class imbalance in medical imaging datasets hinders the development of reliable AI diagnostic systems.
  • Multi-sensor and multi-modal data integration (e.g., CT, MRI, PET) offers potential to address data scarcity for rare diseases.

Purpose of the Study:

  • To critically review and categorize strategies for mitigating class imbalance in AI medical imaging.
  • To explore the role of multi-sensor fusion in enhancing minority class representation.
  • To evaluate the clinical viability and compliance of different imbalance-handling techniques.

Main Methods:

  • Categorization of strategies into data-centric (resampling, synthesis) and model-centric (loss functions, transfer learning, ensembles).
  • Analysis of multi-sensor feature-level and decision-level fusion paradigms.
Keywords:
class imbalanceclinical AI deploymentdata augmentationensemble learningmedical image classificationmulti-modal imagingmulti-sensor fusion

Related Experiment Videos

Last Updated: Jul 1, 2026

Design and Analysis for Fall Detection System Simplification
08:05

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

  • Evaluation of methods based on clinical viability, compliance, and emerging trends like federated learning and explainable AI.

    • Main Results:

    • Data-centric strategies improved minority class recall by 12-35% in imbalanced datasets (≥10:1 ratio).
    • Model-centric strategies achieved an average AUC improvement of 0.08-0.21 in multi-sensor tasks.
    • Multi-sensor fusion inherently enriches representations for underrepresented classes.

    Conclusions:

    • Synergistic approaches combining imbalance-aware learning with multi-sensor fusion, federated learning, and explainable AI are essential.
    • These integrated strategies pave the way for robust, equitable, and clinically deployable AI diagnostic tools.
    • Addressing class imbalance is key to unlocking the full potential of AI in medical diagnostics.