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

Trends of antihypertensive agents in patients with hypertension and coronary artery disease in a tertiary hospital of China.

International journal of clinical pharmacy·2020
Same author

Assessing the association between dipeptidyl peptidase-4 inhibitors use and celiac disease through drug adverse event reporting.

Therapeutic advances in chronic disease·2020
Same author

Mannose receptor (CD206) activation in tumor-associated macrophages enhances adaptive and innate antitumor immune responses.

Science translational medicine·2020
Same author

Pharmacologic Prophylaxis of Stress Ulcer in Non-ICU Patients: A Systematic Review and Network Meta-analysis of Randomized Controlled Trials.

Clinical therapeutics·2020
Same author

Integrative Analysis of Long Non-Coding RNAs (lncRNAs), miRNAs, and mRNA-Associated ceRNA Network in Lung Tissue of Aging Mice and Changes After Treatment with Codonopsis pilosula.

Medical science monitor : international medical journal of experimental and clinical research·2020
Same author

Level of N6-Methyladenosine in Peripheral Blood RNA: A Novel Predictive Biomarker for Gastric Cancer.

Clinical chemistry·2020

Related Experiment Video

Updated: Jul 31, 2025

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

1.9K

A devised thyroid segmentation with multi-stage modification based on Super-pixel U-Net under insufficient data.

Yifei Chen1, Dandan Li1, Xin Zhang1

  • 1Department of Control Science and Engineering, Harbin Institute of Technology, Harbin, 150001 China.

Ultrasound in Medicine & Biology
|May 3, 2023
PubMed
Summary

This study introduces a Super-pixel U-Net for improved thyroid ultrasound image segmentation, overcoming challenges with non-thyroid regions and limited data. The novel method significantly enhances segmentation accuracy and shape similarity for better diagnostic insights.

Keywords:
Image segmentationMulti-stage modificationSuper-pixel U-NetThyroid ultrasound images

More Related Videos

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

2.8K
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.6K

Related Experiment Videos

Last Updated: Jul 31, 2025

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

1.9K
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

2.8K
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.6K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computer Vision

Background:

  • Deep learning for medical image segmentation is advancing rapidly.
  • Segmenting thyroid ultrasound images presents challenges due to non-thyroidal regions and insufficient data.
  • Existing deep learning models struggle to achieve optimal results in thyroid image segmentation.

Purpose of the Study:

  • To develop an improved deep learning model for precise thyroid ultrasound image segmentation.
  • To address the limitations of existing methods in handling non-thyroidal regions and data scarcity.
  • To enhance the accuracy and reliability of automated thyroid segmentation.

Main Methods:

  • A novel Super-pixel U-Net architecture was designed by incorporating a supplementary path into the standard U-Net.
  • A multi-stage modification involving boundary segmentation, boundary repair, and auxiliary segmentation was implemented.
  • The method utilizes U-Net for initial rough segmentation, followed by refinement and precise assistance from the Super-pixel U-Net.

Main Results:

  • The proposed Super-pixel U-Net achieved a high F1 Score of 0.9161 and IoU of 0.9279.
  • Superior performance was observed in shape similarity metrics, including convexity (0.9395), ratio (0.9109), compactness (0.8976), eccentricity (0.9448), and rectangularity (0.9289).
  • The average area estimation indicator reached 0.8857, demonstrating accurate size assessment.

Conclusions:

  • The Super-pixel U-Net demonstrates superior performance in thyroid ultrasound image segmentation.
  • The multi-stage modification effectively improves segmentation accuracy and robustness.
  • The proposed method offers a promising advancement for automated thyroid analysis in medical imaging.