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

Experimental Validation and Bioinformatics Analysis Elucidate the Role of MTDH-Mediated PTEN Ubiquitination and Degradation in Podocyte Injury in Diabetic Kidney Disease.

Human mutation·2026
Same author

Gradient-based rigid motion correction in CBCT via Lie algebra-constrained registration.

Physics in medicine and biology·2026
Same author

BrainUMA: A Unified multi-atlas learning framework for brain disorders diagnosis.

Medical & biological engineering & computing·2026
Same author

C[Formula: see text]Net: A co-occurrence and consistency-aware framework for structured multi-label fundus diagnosis.

Medical & biological engineering & computing·2026
Same author

Size- and Time-Dependent Impacts of Polyvinyl Chloride Microplastics on Turbot (<i>Scophthalmus maximus</i> L.): Intestinal Tolerance, Hepatic Injury, and Intestinal Microbiota Dysbiosis.

Toxics·2026
Same author

From Ethnopharmacology to Drug Discovery: The Therapeutic Potential of Anisomeles indica.

Phytochemical analysis : PCA·2026

Related Experiment Video

Updated: Jan 10, 2026

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

3.3K

Adaptive mix for semi-supervised medical image segmentation.

Zhiqiang Shen1, Peng Cao2, Junming Su1

  • 1School of Computer Science and Engineering, Northeastern University, Shenyang, 110819, China; Key Laboratory of Intelligent Computing in Medical Image, Ministry of Education, Shenyang, 110819, China.

Medical Image Analysis
|November 22, 2025
PubMed
Summary
This summary is machine-generated.

Adaptive Mix (AdaMix) improves semi-supervised learning by adaptively adjusting image perturbations during training. This self-paced approach enhances consistency regularization for better model performance in medical image segmentation tasks.

Keywords:
Medical image segmentationMix-upSelf-paced learningSemi-supervised learning

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

727
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

43.4K

Related Experiment Videos

Last Updated: Jan 10, 2026

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

3.3K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

727
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

43.4K

Area of Science:

  • * Machine Learning
  • * Computer Vision
  • * Medical Image Analysis

Background:

  • * Mix-up is crucial for semi-supervised learning (SSL) consistency regularization, generating perturbed samples for pseudo-supervision.
  • * Current mix-up methods (random or fixed rules) limit effectiveness due to uncontrolled or trivial perturbations.
  • * Adaptive perturbation is needed to optimize SSL performance.

Purpose of the Study:

  • * To investigate adaptive image mix-up perturbation during training for SSL.
  • * To propose an Adaptive Mix (AdaMix) algorithm using a self-paced learning strategy.
  • * To develop and evaluate AdaMix-based frameworks for semi-supervised medical image segmentation.

Main Methods:

  • * Introduced AdaMix, a self-paced learning algorithm for adaptive image mix-up.
  • * Implemented a self-paced curriculum to control perturbation difficulty based on model learning state.
  • * Developed three frameworks: AdaMix-ST, AdaMix-MT, and AdaMix-CT for medical image segmentation.

Main Results:

  • * AdaMix frameworks achieved superior performance on 2D and 3D medical image segmentation tasks across three datasets.
  • * AdaMix-CT demonstrated significant improvements: 2.62% in Dice Similarity Coefficient and 48.25% in Average Surface Distance on the ACDC dataset (10% labeled data).
  • * Dynamically adjusted mix-up perturbation enhanced consistency regularization effectiveness.

Conclusions:

  • * AdaMix enables effective adaptive perturbation for semi-supervised learning.
  • * The proposed self-paced approach optimizes mix-up for improved medical image segmentation.
  • * AdaMix offers a promising direction for enhancing SSL consistency regularization.