Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jan 29, 2026

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

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

Published on: December 15, 2023

Domain Shift in Breast DCE-MRI Tumor Segmentation: A Balanced LoCoCV Study on the MAMA-MIA Dataset.

Munid Alanazi¹, Bader Alsharif²

¹Business Informatics Department, College of Business, King Khalid University, Abha 61421, Saudi Arabia.

Diagnostics (Basel, Switzerland)

|January 28, 2026

Summary

Related Concept Videos

Imaging Studies I: CT and MRI

Imaging Studies I: CT and MRI

Introduction: MRI and CT scans are crucial advancements in medical imaging techniques, playing a vital role in diagnosing conditions related to the gastrointestinal (GI) system. Each scan serves distinct purposes, targets specific areas, and requires unique nursing duties.
Description of the Procedures
Computed Tomography (CT) scan:
Computed Tomography (CT) scans use X-ray technology to generate detailed images of bones, organs, and tissues. During the scan, the patient lies on a moving table...

Balancing Redox Equations

Balancing Redox Equations

Electrochemistry is the science involved in the interconversion of electrical and chemical reactions. Such reactions are called reduction-oxidation, or redox reactions. These important reactions are defined by changes in oxidation states for one or more reactant elements and include a subset of reactions involving the transfer of electrons between reactant species. Electrochemistry as a field has evolved to yield sufficient insights on the fundamental principles of redox chemistry and multiple...

Conservation of Protein Domains Over Different Proteins

Conservation of Protein Domains Over Different Proteins

Protein domains are small structurally independent units that are part of a single amino acid chain. Although these domains are often structurally independent, they may rely on synergistic effects to perform their functions as part of a larger protein. Protein domains may be conserved within the same organism, as well as across different organisms.
A limited set of protein domains often duplicate and recombine during evolution. These domains can be organized in different combinations to...

Equilibrium and Balance

Equilibrium and Balance

The inner ear assumes dual functionalities of auditory perception and equilibrium maintenance. The vestibule is the organ responsible for balance. This organ contains mechanoreceptors, specifically hair cells, endowed with stereocilia, which aid in deciphering information regarding the position and motion of our heads. Two intrinsic components, the utricle and saccule, help perceive head position, while the semicircular canals track head movement. Neurological messages initiated in the...

Energy Balance

Energy Balance

The human body gets energy from the three macronutrients: carbohydrates, proteins, and fats. Energy is released when the chemical bonds in the organic compounds present in the food are broken down. The energy content of food is measured in kilocalories (kcal), defined as the amount of heat required to raise the temperature of one kilogram of water by one degree Celsius. This value is determined by measuring the temperature change of the water surrounding a calorimeter after the complete...

Membrane Domains

Membrane Domains

The membrane domains concentrate specific lipids and proteins at one place within the membrane, which helps in cell signaling, adhesion, and other critical cellular processes. These domains can differ in size, composition, function, and lifespan.
Protein Domains
The membrane comprises a group of distinct proteins responsible for carrying out a cell's specific function. For example, the plasma membrane of the human sperm, or a single germ cell, contains a unique set of proteins in the...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Real-Time American Sign Language Interpretation Using Deep Learning and Keypoint Tracking.

Sensors (Basel, Switzerland)·2025

Same author

Stacking Ensemble Deep Learning for Real-Time Intrusion Detection in IoMT Environments.

Sensors (Basel, Switzerland)·2025

Same author

Enhancing Cybersecurity in Healthcare: Evaluating Ensemble Learning Models for Intrusion Detection in the Internet of Medical Things.

Sensors (Basel, Switzerland)·2024

Same author

Deep Learning Technology to Recognize American Sign Language Alphabet.

Sensors (Basel, Switzerland)·2023

Deep learning models for breast MRI segmentation struggle with performance drops at new hospitals due to domain shift. Robust multi-center training and validation are essential for reliable clinical deployment.

Area of Science:

Medical Imaging
Artificial Intelligence in Medicine
Oncology

Background:

Accurate breast tumor segmentation in dynamic contrast-enhanced MRI (DCE-MRI) is vital for cancer treatment and monitoring.
Deep learning models often exhibit performance degradation when applied to data from different hospitals due to variations in hardware, protocols, and patient populations (center-related domain shift).

Purpose of the Study:

To investigate the impact of center-related domain shift on automated breast DCE-MRI tumor segmentation using the multi-center MAMA-MIA dataset.
To evaluate the effectiveness of different cross-validation strategies in assessing model generalizability across institutions.

Main Methods:

Trained a 3D U-Net model for primary tumor segmentation.
Evaluated performance using two settings: a random patient-wise split (in-distribution) and a balanced leave-one-center-out cross-validation (LoCoCV) protocol (out-of-distribution).

Keywords:

DCE-MRI MAMA-MIA dataset U-Net breast cancer deep learning domain shift leave-one-center-out cross-validation (LoCoCV)multi-center learning tumor segmentation

More Related Videos

Terahertz Imaging and Characterization Protocol for Freshly Excised Breast Cancer Tumors

Terahertz Imaging and Characterization Protocol for Freshly Excised Breast Cancer Tumors

Published on: April 5, 2020

Orthotopic Transplantation of Breast Tumors as Preclinical Models for Breast Cancer

Orthotopic Transplantation of Breast Tumors as Preclinical Models for Breast Cancer

Published on: May 18, 2020

Related Experiment Videos

Last Updated: Jan 29, 2026

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

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

Published on: December 15, 2023

Terahertz Imaging and Characterization Protocol for Freshly Excised Breast Cancer Tumors

Terahertz Imaging and Characterization Protocol for Freshly Excised Breast Cancer Tumors

Published on: April 5, 2020

Orthotopic Transplantation of Breast Tumors as Preclinical Models for Breast Cancer

Orthotopic Transplantation of Breast Tumors as Preclinical Models for Breast Cancer

Published on: May 18, 2020

Assessed segmentation accuracy using Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD95), sensitivity, and specificity.

Main Results:

On the mixed-center random split, the model achieved a mean Dice of ~0.68 and HD95 of ~19.7 mm, indicating good performance when training and test distributions match.
Under balanced LoCoCV, performance significantly degraded with a mean Dice of ~0.45 and HD95 of ~41 mm on unseen centers, demonstrating substantial accuracy loss.
Cross-center validation revealed increased boundary errors and reduced segmentation reliability when models are applied to new institutions.

Conclusions:

Models performing well on mixed-center data can experience significant accuracy loss on unseen institutions.
The balanced LoCoCV protocol effectively highlights the out-of-distribution penalty caused by center-related domain shift.
Emphasizes the critical need for robust multi-center training strategies and explicit cross-center validation for clinical deployment of breast DCE-MRI segmentation models.