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

Standardized Reporting of Cardiac Magnetic Resonance Examinations in Children With Cardiac Diseases and Adults With Congenital Heart Disease: A Scientific Statement From the Association for European Pediatric and Congenital Cardiology (AEPC) and the International Society for Magnetic Resonance in Medicine (ISMRM).

Journal of magnetic resonance imaging : JMRI·2026
Same author

Estimating Acrylamide and 5-Hydroxymethylfurfural Levels in Crackers Using Computer Vision: Effects on Consumer Acceptance.

Foods (Basel, Switzerland)·2026
Same author

Multiparametric Free-Breathing 3D Whole-Heart Cardiac MR for Anatomical Bright- and Black-Blood Imaging With Co-Registered <math><semantics><mrow><msub><mrow><mi>T</mi></mrow> <mrow><mn>1</mn></mrow></msub> <mo>/</mo> <msub><mrow><mi>T</mi></mrow> <mrow><mn>2</mn></mrow></msub></mrow> <annotation>$$ {T}_1/{T}_2 $$</annotation></semantics></math> Myocardial Tissue Mapping at <math><semantics><mrow><mn>0</mn> <mo>.</mo> <mn>55</mn></mrow> <annotation>$$ 0.55 $$</annotation></semantics></math> T.

NMR in biomedicine·2026
Same author

Single-Breathhold 3D MR Elastography in the Liver, With Simultaneous R2* and PDFF Mapping.

Magnetic resonance in medicine·2026
Same author

Molecular Imaging of Coronary Plaque Vulnerability Using 18F-Fluorocholine PET-MRI in Patients with Coronary Artery Disease: Validation with Optical Coherence Tomography.

Journal of clinical medicine·2025
Same author

3D Whole-Heart Joint T<sub>1</sub>/T<sub>1ρ</sub> Mapping and Water-Fat Imaging on a Clinical 0.55-T Low-Field Scanner.

NMR in biomedicine·2025

Related Experiment Video

Updated: Sep 12, 2025

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

2.9K

In Silico Digital Breast Tomosynthesis Dataset for the Comparative Analysis of Deep Learning Models in Tumor

Cristina Alfaro Vergara1,2, Nicolás Araya Caro3, Domingo Mery Quiroz3,4

  • 1Department of Medical Technology, Faculty of Health Sciences, Universidad de Tarapacá, Arica, Chile. calfarov@academicos.uta.cl.

Journal of Imaging Informatics in Medicine
|August 5, 2025
PubMed
Summary

This study shows that computer-generated (in silico) digital breast tomosynthesis (DBT) data can effectively train deep learning models for breast tumor segmentation, addressing the scarcity of real-world data.

Keywords:
Breast tumor segmentationDeep learning U-NetDigital breast tomosynthesis (DBT)Hybrid trainingIn silico data

More Related Videos

Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

9.3K
Using Computer-based Image Analysis to Improve Quantification of Lung Metastasis in the 4T1 Breast Cancer Model
08:32

Using Computer-based Image Analysis to Improve Quantification of Lung Metastasis in the 4T1 Breast Cancer Model

Published on: October 2, 2020

6.4K

Related Experiment Videos

Last Updated: Sep 12, 2025

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

2.9K
Deep Learning-Based Segmentation of Cryo-Electron Tomograms
10:25

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

9.3K
Using Computer-based Image Analysis to Improve Quantification of Lung Metastasis in the 4T1 Breast Cancer Model
08:32

Using Computer-based Image Analysis to Improve Quantification of Lung Metastasis in the 4T1 Breast Cancer Model

Published on: October 2, 2020

6.4K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computational Biology

Background:

  • Limited availability of digital breast tomosynthesis (DBT) datasets hinders the development of deep learning (DL) models for breast tumor segmentation.
  • In silico data generation offers a potential solution to augment real-world datasets.

Purpose of the Study:

  • To assess the feasibility of using in silico-generated DBT data for training DL models for breast tumor segmentation.
  • To compare the performance of various DL architectures trained on in silico data.

Main Methods:

  • Trained 13 DL models (U-Net, FCN, DeepLabv3, DeepLabv3+) on 230 in silico 2D ROIs.
  • Models were trained from scratch or fine-tuned using COCO-pretrained weights (ResNet50/101).
  • Evaluated performance using F1-score, IoU, precision, and recall.

Main Results:

  • U-Net (trained from scratch) and DeepLabv3+ (fine-tuned with ResNet50) showed the best performance (F1-scores ~82-85%, IoUs ~78-84%).
  • No statistically significant differences were observed between top-performing models.
  • Retraining U-Net on a hybrid in silico/real-world dataset yielded a promising 79% F1-score.

Conclusions:

  • In silico DBT data is a viable complementary resource for training DL models in data-limited scenarios.
  • This study provides evidence for integrating computationally generated data into AI-based DBT tumor segmentation research.
  • Further research can leverage in silico data to improve the robustness and accessibility of breast cancer detection tools.