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

Calcium/calmodulin-dependent protein kinase II links ER stress with Fas and mitochondrial apoptosis pathways.

The Journal of clinical investigation·2009
Same author

Cripto-1 overexpression is involved in the tumorigenesis of nasopharyngeal carcinoma.

BMC cancer·2009
Same author

Range of motion and orientation of the lumbar facet joints in vivo.

Spine·2009
Same author

[Silencing of COX-2 in nasopharyngeal carcinoma cells with a shRNAmir lentivirus vector].

Nan fang yi ke da xue xue bao = Journal of Southern Medical University·2009
Same author

The risk of melamine-induced nephrolithiasis in young children starts at a lower intake level than recommended by the WHO.

Pediatric nephrology (Berlin, Germany)·2009
Same author

Adult scoliosis in patients over sixty-five years of age: outcomes of operative versus nonoperative treatment at a minimum two-year follow-up.

Spine·2009

Related Experiment Video

Updated: Sep 24, 2025

High-resolution Functional Magnetic Resonance Imaging Methods for Human Midbrain
10:06

High-resolution Functional Magnetic Resonance Imaging Methods for Human Midbrain

Published on: May 10, 2012

13.0K

Multi-Scale Self-Supervised Learning for Multi-Site Pediatric Brain MR Image Segmentation with Motion/Gibbs

Yue Sun1,2, Kun Gao2, Weili Lin2

  • 1Department of Shandong Provincial Key Laboratory of Network based Intelligent Computing, University of Jinan, Jinan 250022, China.

Machine Learning in Medical Imaging. MLMI (Workshop)
|May 9, 2022
PubMed
Summary

This study introduces a multi-scale self-supervised learning (M-SSL) framework for accurate pediatric brain MRI segmentation. The method effectively handles artifacts and multi-site data variations, improving early brain development characterization.

Keywords:
Deep LearningMotion/Gibbs ArtifactsMulti-Site IssuePediatric Brain Segmentation

More Related Videos

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

2.9K
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

42.9K

Related Experiment Videos

Last Updated: Sep 24, 2025

High-resolution Functional Magnetic Resonance Imaging Methods for Human Midbrain
10:06

High-resolution Functional Magnetic Resonance Imaging Methods for Human Midbrain

Published on: May 10, 2012

13.0K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

2.9K
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

42.9K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Neuroscience

Background:

  • Accurate segmentation of pediatric brain MRI is crucial for understanding early brain development.
  • Challenges include imaging artifacts (motion, Gibbs) and multi-site data variability (domain shift).
  • Existing methods struggle with these complexities in large-scale datasets.

Purpose of the Study:

  • To develop a robust framework for accurate brain tissue segmentation in multi-site pediatric MRI data.
  • To address challenges posed by imaging artifacts and domain shift.
  • To improve the characterization of early brain development through enhanced segmentation.

Main Methods:

  • A multi-scale self-supervised learning (M-SSL) framework was proposed.
  • Coarse segmentation on downsampled images provided global anatomic guidance.
  • Fine segmentation on original images was refined using this guidance.
  • An iterative self-supervised strategy trained site-specific models to mitigate multi-site issues.

Main Results:

  • The M-SSL method demonstrated superior performance compared to state-of-the-art approaches.
  • Experiments were conducted on the iSeg2019 challenge dataset, featuring real artifacts and multi-site pediatric brain MR images.
  • The framework successfully segmented brain tissues with high accuracy despite data complexities.

Conclusions:

  • The proposed M-SSL framework offers an effective solution for segmenting pediatric brain MR images with artifacts and multi-site variations.
  • This advancement facilitates more accurate characterization of early brain development.
  • The method shows significant potential for clinical and research applications in pediatric neuroimaging.