Using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images.

Gustav Müller-Franzes, Luisa Huck, Soroosh Tayebi Arasteh, Firas Khader, Tianyu Han, Volkmar Schulz, Ebba Dethlefsen, Jakob Nikolas Kather, Sven Nebelung, Teresa Nolte, Christiane Kuhl, Daniel Truhn

May 2023 Radiology

Synopsis of Social media discussions

The discussions emphasize the transformative potential of the research, with comments like 'game changer if/when it pans out' and 'tremendous promise,' which indicate excitement about future possibilities. The tone is optimistic, and the references to GANs and low-dose imaging suggest a focus on technological advances and their significance for patient care, fostering a sense of deep engagement and anticipation among participants.

A
Agreement
 Moderate agreement

Most discussions express strong support or enthusiasm for the research, using words like 'tremendous promise,' 'game changer,' and 'proof of concept,' indicating a high level of agreement on its significance.

I
Interest
 High level of interest

Posts demonstrate high curiosity and relevance, with many highlighting the potential to reduce contrast agents and mentioning specific techniques like GANs and low-dose imaging, reflecting keen interest.

E
Engagement
 High engagement

Several comments delve into the technical implications, such as GAN recovery and lesion detection, showing deep engagement with the research's methods and potential applications.

I
Impact
 Moderate level of impact

The discussions suggest that the publication could have a meaningful impact on medical imaging practices, though some posts mention it as a promising proof of concept rather than an immediate breakthrough, indicating perceived high but cautious impact.

Social Mentions

YouTube

2 Videos

Twitter

26 Posts

News

2 Articles

Metrics

Video Views

401

Total Likes

59

Extended Reach

174,500

Social Features

30

Timeline: Posts about article

Top Social Media Posts

Posts referencing the article

Reducing Contrast Agent Use in Breast MRI with Machine Learning Techniques

Reducing Contrast Agent Use in Breast MRI with Machine Learning Techniques

Interview discussing how machine learning can generate synthetic contrast images in breast MRI, reducing the need for contrast agents, with insights from Dr. Daniel Truhn.

EUSOBI Official YouTube Channel

May 22, 2023

308 views

Machine Learning Reduces Contrast Agent Need in Breast MRI

Machine Learning Reduces Contrast Agent Need in Breast MRI

Interview discusses a study on using neural networks to generate synthetic breast MRI images, potentially minimizing contrast agent use. Daniel Truhn shares insights on the technology and future applications.

EUSOBI Official YouTube Channel

May 22, 2023

93 views

  • D. Pinto dos Santos
    @pintodrad (Twitter)

    RT @EUSOBIyc:
    view full post

    May 19, 2023

    1

  • EUSOBI
    @EUSOBIyc (Twitter)


    view full post

    May 19, 2023

    1

    1

  • Gustav Müller-Franzes
    @FranzesGustav (Twitter)

    RT @EUSOBIyc:
    view full post

    May 19, 2023

    3

  • Matthias Dietzel, Prof. Dr., MD, MHBA
    @mad_rad_007 (Twitter)

    RT @EUSOBIyc:
    view full post

    May 15, 2023

    3

  • D. Pinto dos Santos
    @pintodrad (Twitter)

    RT @EUSOBIyc:
    view full post

    May 12, 2023

    3

  • EUSOBI
    @EUSOBIyc (Twitter)


    view full post

    May 12, 2023

    6

    3

  • Pilar Manchón
    @pilarmanchon_rx (Twitter)

    RT @EUSOBIyc:
    view full post

    April 1, 2023

    2

  • الدكتورة فاتنه محمد الطحان
    @Fatina_t (Twitter)

    RT @EUSOBIyc:
    view full post

    March 31, 2023

    2

  • EUSOBI
    @EUSOBIyc (Twitter)


    view full post

    March 31, 2023

    2

    2

  • Jean Seely, MD
    @JeanSeely (Twitter)

    This proof of concept study using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images offers tremendous promise for Breast MRI. https://t.co/AJO6zIu59b
    view full post

    March 25, 2023

    4

  • Salvador Pedraza
    @salvasapedraza (Twitter)

    Using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images https://t.co/lCcOIj5YM5 https://t.co/vUWIK5Vu6E
    view full post

    March 25, 2023

    1

  • Salvador Pedraza
    @salvasapedraza (Twitter)

    Using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images. Gustav Müller-Franzes https://t.co/CFyUSR3mjE https://t.co/Tl3YGlpWNg
    view full post

    March 25, 2023

  • Zekun Jiang
    @zekun_jiang (Twitter)

    RT @radiology_rsna: Using machine learning to reduce the need for contrast agents in breast MRI through synthetic images​: learn more in th…
    view full post

    March 24, 2023

    1

  • Dr. Paula Gordon
    @DrPaulaGordon (Twitter)

    This will be a game changer if/when it pans out. https://t.co/aSqapchAxX https://t.co/55dlVcb5rt
    view full post

    March 23, 2023

    4

  • Radiology
    @radiology_rsna (Twitter)

    Using machine learning to reduce the need for contrast agents in breast MRI through synthetic images​: learn more in this editorial by Dr. Bahl of @MGHImaging. https://t.co/qDmSsykbi2 https://t.co/3iqkG6Oukl
    view full post

    March 23, 2023

    4

    1

  • 本田茉也　Maya Honda
    @mayahf0217 (Twitter)

    RT @radiology_rsna: Using a training set of standardized contrast-enhanced breast MRI scans, GANs recovered image appearance and conspicuit…
    view full post

    March 22, 2023

    4

  • BCRS
    @BCRadSoc (Twitter)

    RT @radiology_rsna: Using a training set of standardized contrast-enhanced breast MRI scans, GANs recovered image appearance and conspicuit…
    view full post

    March 22, 2023

    4

  • Haidara Almansour, MD, M.Eng
    @HalmansourMD (Twitter)

    RT @DrLindaMoy: #RADIOLOGY AI generated low dose breast MRI https://t.co/d0u2ilQsYR @radiology_rsna @ProfKuhl #breastradiology #breastmri…
    view full post

    March 22, 2023

    2

  • Radiology
    @radiology_rsna (Twitter)

    Using a training set of standardized contrast-enhanced breast MRI scans, GANs recovered image appearance and conspicuity of enhancing lesions from simulated low-contrast images, as verified on independent test sets. https://t.co/2hPbOHPNi6 https://t.co/3z2wTrmcgN
    view full post

    March 22, 2023

    13

    4

  • LN Tanenbaum MD FACR
    @nuromri (Twitter)

    Using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images | Radiology https://t.co/4RvuzSbtGW
    view full post

    March 22, 2023

    1

  • Durgesh Dwivedi ✍
    @durgeshdwivedi (Twitter)

    RT @DrLindaMoy: #RADIOLOGY AI generated low dose breast MRI https://t.co/d0u2ilQsYR @radiology_rsna @ProfKuhl #breastradiology #breastmri…
    view full post

    March 21, 2023

    2

  • Linda Moy
    @DrLindaMoy (Twitter)

    #RADIOLOGY AI generated low dose breast MRI https://t.co/d0u2ilQsYR @radiology_rsna @ProfKuhl #breastradiology #breastmri #mri https://t.co/HpDUgrFGBy
    view full post

    March 21, 2023

    17

    2

  • Dr M. Mahesh (ಮಹೇಶ್) (he/him/his)
    @mmahesh1 (Twitter)

    Interesting proof of concept: Aiming to reduce Gd contrast dose by enhancing low-dose contrast breast MRI images w Generative Adversarial Networks (GAN) Probably one can work reduce Iodine contrast in contrast CT too! https://t.co/HbIJQscKYK https://t.co/FVIDbOdXr6
    view full post

    March 21, 2023

    4

  • Tim Leiner
    @MLandDL_papers (Twitter)

    Using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images https://t.co/oMTuursIM9
    view full post

    March 21, 2023

  • NK papers
    @NK_papers (Twitter)

    Using Machine Learning to Reduce the Need for Contrast Agents in Breast MRI through Synthetic Images by Gustav Müller-Franzes https://t.co/CASm6osJiT
    view full post

    March 21, 2023

  • Egutz
    @Egutz_ (Twitter)

    Using Machine Learning to Reduce the Need for Contrast Agents in ... - RSNA Publications Online https://t.co/ZibUUBiXev
    view full post

    March 21, 2023

Abstract Synopsis

  • The study investigated whether GANs can create synthetic contrast-enhanced breast MRI images from unenhanced or low-contrast images, aiming to reduce the use of contrast agents.
  • Results showed that radiologists struggled to distinguish real from synthetic images, with accuracy around 52-61%, indicating high similarity between real and generated images, especially using approach A.
  • The findings suggest that GANs could be a promising tool to produce realistic MRI images, potentially decreasing the need for contrast agents in breast imaging.