Presentations & organization

Talks & community.

My conference presentations, invited talks, tutorials, and community roles.

Talks, tutorials & posters

  1. Oct 2026

    Tutorial · ML in PL · Warsaw, Poland

    Introduction to Physics Neural Surrogates for Engineering and Healthcare

    A hands-on tutorial with Julian Suk on geometric learning, neural fields, neural operators, and physical inductive biases.

  2. Jun 2026

    Contributed talk · 21st International Symposium on Biomechanics in Vascular Biology and Cardiovascular Disease · Geneva, Switzerland

    Equivariance drives data-efficiency in deep-learning-based estimation of hemodynamics in cardiovascular diseases

    Presented in the session on advanced imaging for biomechanical forces, wall structure and function at the Shear Stress Symposium.

  3. Apr 2026

    Invited talk · Medis headquarters · Leiden, the Netherlands

    Neural Surrogates for Hemodynamics Estimation in Cardiovascular Diseases

    Invited presentation on neural surrogates for cardiovascular hemodynamics.

  4. Jan 2026

    Talk · Best Talk Award · MISP² · Leiden, the Netherlands

    Neural Surrogates for Hemodynamics Estimation in Cardiovascular Diseases

    Geometric neural surrogates for blood-flow estimation, with an emphasis on data efficiency and physical structure.

  5. Nov 2025

    Talk · Cardiovascular Biomechanics NL (CVB-NL) · TU Delft, the Netherlands

    AI-driven surrogate models for flow estimation in abdominal aortic aneurysms

    Presented at the Dutch cardiovascular biomechanics meeting, hosted by the Department of Mechanical Engineering at TU Delft.

  6. Oct 2025

    Masterclass · Dutch Inverse Problems Meeting · University of Twente, Enschede, the Netherlands

    Neural surrogates in medicine

    A contribution on neural surrogates in medicine to the masterclass on deep learning for medical image processing at the fifth Dutch Inverse Problems meeting.

  7. Sep 2025

    Talk & poster · STACOM at MICCAI · Daejeon, South Korea

    Active Learning for Deep Learning-Based Hemodynamic Parameter Estimation

    Selecting informative simulations to train hemodynamic surrogates with less reference data.

  8. Jun 2025

    Invited talk · Innovation in vascular surgery · Meander Medical Centre, the Netherlands

    Neural surrogates for blood flow dynamics

    Invited presentation at a conference on innovation in vascular surgery.

  9. May 2025

    Talk · GHOST Day · Poznań, Poland

    Geometric deep learning for blood flow modelling in cardiovascular diseases

    An introduction to learning blood-flow models directly from vascular geometry.

  10. Oct 2024

    Poster · Best Poster Award · STACOM at MICCAI · Marrakesh, Morocco

    Global Control for Local SO(3)-Equivariant Scale-Invariant Vessel Segmentation

    Combining local symmetry-aware segmentation with global control over the vascular model.

  11. Oct 2023

    Co-authored research · ML in PL · Warsaw, Poland

    Artificial Intelligence for non-invasive cardiac diagnostics

    CenterlinePointNet++ research presented by Tomasz Konopczyński in a contributed talk on non-invasive cardiac diagnostics.

  12. Oct 2023

    Poster · MICCAI · Vancouver, Canada

    CenterlinePointNet++: A New Point Cloud Based Architecture for Coronary Artery Pressure Drop and vFFR Estimation

    Poster on learning coronary pressure drop and virtual fractional flow reserve from vessel geometry.

  13. Apr 2023

    Talk · GHOST Day · Poznań, Poland

    Applications of neural networks for point clouds in vascular medical image processing

    Point-cloud networks for medical imaging, including vessel labeling and vascular geometry analysis.

  14. Nov 2022

    Poster · GeoMedIA workshop · Amsterdam, the Netherlands

    Eigenvector Grouping for Point Cloud Vessel Labeling

    Geometric deep learning for coronary vessel labeling using point-cloud neighborhoods shaped to vascular anatomy.

  15. Nov 2022

    Poster · co-author · ML in PL · Warsaw, Poland

    Coronary Ostia Localization Using Residual U-Net with Heatmap Matching and 3D DSNT

    Joint work with Miłosz Gajowczyk and Tomasz Konopczyński on localizing coronary landmarks in CT images.

  16. Sep 2022

    Workshop presentation · MLMI workshop at MICCAI · Singapore

    Coronary Ostia Localization Using Residual U-Net with Heatmap Matching and 3D DSNT

    Presented the coronary ostia localization paper at the Machine Learning in Medical Imaging workshop.

  17. Nov 2021

    Poster · ML in PL · Online

    Artefact removal from 3D coronary artery segmentation from CT via point cloud neural network

    Point-cloud classification for separating coronary vessels from segmentation artefacts, with Tomasz Konopczyński and Maciej Zięba.

Organization & community

  1. 2026

    LOGML Summer School · Imperial College London, United Kingdom

    Co-organizer

    Co-organized the London Geometry and Machine Learning Summer School as part of the speakers and mentors team.

  2. 202620252024

    ML in PL · Warsaw, Poland

    Association member & conference organizer

    1. 2026 · Call for Contributions team coordinator

      Coordinated the conference's Call for Contributions team.

    2. 2025 · Tutorials coordinator & Call for Contributions team

      Helped organize tutorials and contributed research sessions.

    3. 2024 · Association member & Call for Contributions team

      Joined the association and the conference's Call for Contributions team; co-organized the Utrecht meetup.

  3. 2025

    MISP² · University of Twente, the Netherlands

    General chair

    Chaired the Medical Imaging Symposium for Netherlands-based PhD candidates and postdoctoral researchers on 15 January 2025.

  4. 2023

    SOLVRO · Wrocław University of Science and Technology, Poland

    Machine Learning section leader

    Section member from October 2019 to September 2023 and leader from September 2021. Also represented MSc AI students on the curriculum board in 2022–2023.

  5. 2023

    Charity events

    Volunteer robotics teacher

    Taught LEGO robotics at charitable events between 2020 and 2023.