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

Office Hysteroscopy: Treating Intrauterine Pathology and Avoiding the Operating Room.

JSLS : Journal of the Society of Laparoendoscopic Surgeons·2026
Same author

Effectiveness of a Personalized Digital Exercise and Nutrition Rehabilitation Program in Postoperative Patients With Gastric Cancer: Randomized Controlled Trial.

JMIR mHealth and uHealth·2026
Same author

Prevalence of HER2-ultralow breast cancer in South Korea: a multicenter study by reassessment of HER2-zero cases.

Journal of pathology and translational medicine·2026
Same author

Next generation protein-corrole bio-assemblies provide effective tumoricidal treatment in a metastatic triple-negative breast cancer model.

bioRxiv : the preprint server for biology·2026
Same author

IDH-mutant gliomas arise from glial progenitor cells harboring the initial driver mutation.

Science (New York, N.Y.)·2026
Same author

Hydralazine inhibits cysteamine dioxygenase to treat preeclampsia and senesce glioblastoma.

Science advances·2025

Related Experiment Video

Updated: Jun 25, 2025

A Label-Free Segmentation Approach for Intravital Imaging of Mammary Tumor Microenvironment
10:39

A Label-Free Segmentation Approach for Intravital Imaging of Mammary Tumor Microenvironment

Published on: May 24, 2022

2.3K

MurSS: A Multi-Resolution Selective Segmentation Model for Breast Cancer.

Joonho Lee1, Geongyu Lee1, Tae-Yeong Kwak1

  • 1Deep Bio Inc., Seoul 08380, Republic of Korea.

Bioengineering (Basel, Switzerland)
|May 25, 2024
PubMed
Summary

A new multi-resolution selective segmentation (MurSS) model accurately segments breast cancer lesions in whole-slide images. This AI approach enhances personalized treatment by improving lesion identification and stability.

Keywords:
breast cancermulti-resolutionsegmentationselective segmentation method

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

389
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

2.4K

Related Experiment Videos

Last Updated: Jun 25, 2025

A Label-Free Segmentation Approach for Intravital Imaging of Mammary Tumor Microenvironment
10:39

A Label-Free Segmentation Approach for Intravital Imaging of Mammary Tumor Microenvironment

Published on: May 24, 2022

2.3K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

389
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

2.4K

Area of Science:

  • Digital Pathology
  • Computational Oncology
  • Medical Image Analysis

Background:

  • Accurate segmentation of cancer lesions is crucial for personalized cancer treatment and improving patient outcomes.
  • Hematoxylin and eosin (H&E) stained whole-slide images (WSIs) are standard for cancer diagnosis but require precise analysis.
  • Deep learning models offer potential for automating and enhancing the accuracy of cancer lesion segmentation.

Purpose of the Study:

  • To develop and evaluate a novel multi-resolution selective segmentation (MurSS) model for accurate breast cancer lesion segmentation in H&E stained WSIs.
  • To leverage multi-resolution features and a selective mechanism to improve segmentation performance and stability.
  • To compare the performance of the MurSS model against existing deep learning methods.

Main Methods:

  • The MurSS model was developed using both low- and high-resolution image patches from WSIs.
  • Adaptive instance normalization was employed to integrate multi-resolution features effectively.
  • A selective segmentation method was implemented to automatically reject ambiguous tissue regions, ensuring training stability.
  • The model was trained and validated on The Cancer Genome Atlas breast invasive carcinoma (BRCA) dataset and tested on a separate BRCA dataset from Korea University Medical Center.

Main Results:

  • The MurSS model achieved a pixel-level accuracy of 96.88% (95% CI: 95.97-97.62%) and a mean Intersection over Union (IoU) of 0.7283 (95% CI: 0.6865-0.7640).
  • The model demonstrated superior performance compared to other deep learning models evaluated.
  • MurSS successfully rejected approximately 5% of ambiguous WSI regions, aligning with expert annotations.

Conclusions:

  • The MurSS model provides accurate and stable segmentation of breast cancer lesions from WSIs.
  • Its ability to utilize multi-resolution information and reject ambiguous regions makes it a promising tool for digital pathology and personalized medicine.
  • The MurSS model represents a significant advancement in automated cancer lesion segmentation for improved diagnostic and therapeutic strategies.