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

  1. Home
  3. Research Domains
  5. Biomedical And Clinical Sciences
  7. Oncology And Carcinogenesis
  9. Predictive And Prognostic Markers
  11. Application Of Deep Learning On Mammographies To Discriminate Between Low And High-risk Dcis For Patient Participation In Active Surveillance Trials.
  1. Home
  3. Research Domains
  5. Biomedical And Clinical Sciences
  7. Oncology And Carcinogenesis
  9. Predictive And Prognostic Markers
  11. Application Of Deep Learning On Mammographies To Discriminate Between Low And High-risk Dcis For Patient Participation In Active Surveillance Trials.

Related Experiment Video

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

42.9K

Application of deep learning on mammographies to discriminate between low and high-risk DCIS for patient participation in active surveillance trials.

Sena Alaeikhanehshir1,2, Madelon M Voets1,3, Frederieke H van Duijnhoven2

  • 1Division of Molecular Pathology, the Netherlands Cancer Institute, Amsterdam, Netherlands.

Cancer Imaging : the Official Publication of the International Cancer Imaging Society
|April 4, 2024

View abstract on PubMed

Summary
This summary is machine-generated.

A convolutional neural network (CNN) can differentiate low-risk (grade I/II) Ductal Carcinoma In Situ (DCIS) from high-risk (grade III) DCIS or invasive breast cancer using mammography. This AI tool aids in identifying patients suitable for active surveillance, potentially reducing unnecessary surgeries.

Keywords:
Active surveillanceArtificial intelligenceDCISDCIS grade

More Related Videos

Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
15:48

Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging

Published on: December 15, 2014

22.5K
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

6.8K

Related Experiment Videos

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

42.9K
Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
15:48

Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging

Published on: December 15, 2014

22.5K
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

6.8K

Area of Science:

  • Radiology
  • Artificial Intelligence
  • Oncology

Background:

  • Ductal Carcinoma In Situ (DCIS) has varying progression risks; active surveillance is being trialed for low-risk cases (grade I/II).
  • Accurate DCIS grading on mammography is crucial for patient selection in active surveillance trials, especially when pre-surgery biopsies guide eligibility.
  • Current methods for assessing DCIS grade on mammography require improvement to support clinical decision-making in active surveillance protocols.

Purpose of the Study:

  • To evaluate a convolutional neural network (CNN) for distinguishing high-risk (grade III) DCIS and/or Invasive Breast Cancer (IBC) from low-risk (grade I/II) DCIS using mammographic features.
  • To assess the CNN's utility as a decision support tool for excluding high-risk patients from active surveillance programs.
  • To determine the performance of a deep learning model in classifying DCIS risk based on imaging characteristics.
Deep learning
Invasive breast cancer

Main Methods:

  • A retrospective study included 464 patients with pre-surgery biopsy-diagnosed DCIS.
  • A U-Net based CNN was trained on 681 mammograms (80% of cases) and validated on 173 mammograms (20% of cases).
  • Classification performance was measured using Area Under the Curve (AUC) and predictive values for differentiating DCIS risk categories.

Main Results:

  • The CNN achieved an AUC of 0.72 for classifying DCIS as high-risk (grade III), with a Negative Predictive Value (NPV) of 0.91.
  • For distinguishing high-risk DCIS and/or occult IBC from low-risk DCIS, the CNN achieved an AUC of 0.76, with a PPV of 0.80 and NPV of 0.84.
  • The model demonstrated significant capability in discriminating between low-grade and high-grade DCIS on mammography.

Conclusions:

  • The developed CNN effectively discriminates between low-grade (I/II) and high-grade (III) DCIS, with or without co-existing IBC.
  • The high AUC values (0.72 and 0.76) suggest the CNN's potential as a valuable tool in clinical practice.
  • This AI approach can support active surveillance decisions by identifying patients unlikely to have high-risk disease.