iiinova — where your digital self keeps evolving

Reposted by TUM AI in Medicine Lab

Julian Suk @sukjulian.bsky.social · 27/03/2026

One month left 'til @iclr-conf.bsky.social, about time to launch our ✨@gram-org.bsky.social Competititon✨ The theme is geometry x AI4science with a dataset kindly provided by BeyondMath. Deadline: April 22, 2026 (AoE) 🔗 gram-competition.github.io

163

Reposted by TUM AI in Medicine Lab

Marton Szep @martonszep.bsky.social · 10/03/2026

Thrilled to present our paper "Unintended Memorization of Sensitive Information in Fine-Tuned Language Models" at #EACL2026 in Rabat! 🇲🇦 w/ J. Marin Ruiz, G. Kaissis, P. Seidl, R. v. Eisenhart-Rothe, F. Hinterwimmer & @danielrueckert.bsky.social. Read here: arxiv.org/abs/2601.174...

A promotional graphic for an oral presentation at the EACL 2026 conference in Morocco. The background features a sunny, historic Moroccan stone fortress gate with palm trees, a clear blue sky, and decorative geometric tile patterns in the corners. Text in the top left indicates the event is at Palais Des Congres, Rabat, from March 24-29, 2026. A banner across the middle displays the presentation title: "Unintended Memorization of Sensitive Information in Fine-Tuned Language Models." Below the title is a flowchart diagram illustrating how Large Language Models (LLMs) trained on sensitive medical text can inadvertently memorize Personally Identifiable Information (PII), and how a "True-Prefix Attack" can extract a patient's name even when fine-tuned for downstream tasks that do not contain PII. Text at the very bottom reads, "Oral Presentation: March 27 | 11:00 AM | Salle La Palmeraie."

252

Reposted by TUM AI in Medicine Lab

Julian Suk @sukjulian.bsky.social · 27/02/2026

Our latest paper on ML-based flow estimation ✨trained on 4D flow MRI✨ in the carotid arteries is now published (open-access) in Medical Image Analysis. 🔗 www.sciencedirect.com/science/arti...

sciencedirect.com

Physics-informed graph neural networks for flow field estimation in carotid arteries

Hemodynamic quantities are valuable biomedical risk factors for cardiovascular pathology such as atherosclerosis. Non-invasive, in-vivo measurement of…

131

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 02/02/2026

January Thesis Highlights from AIM Lab 🎓 To celebrate their hard work, we want to showcase the excellent research our bachelor and master students produce! This month, Florian Braunmiller finished a great master thesis on Mixture-of-Experts (MoE) architectures for medical imaging. 1/n

Student in front of a research poster on Mixture-of-Experts

110

Reposted by TUM AI in Medicine Lab

ELLIS @ellis.eu · 19/12/2025

🦄 Build Europe’s next AI unicorn - @sprind-de.bsky.social Next Frontier AI initiative backs bold ideas with €125M to create European frontier AI labs. Believe Europe should lead in AI? Join the challenge! 👉 next-frontier.ai

094

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 17/12/2025

Great having Robbie Holland back for a visit at our lab! He’s at AIMI at Stanford University and gave a speech on auto-generated hypotheses in the context of AI4Science. Not only for the lab, but maybe more importantly for our lab‘s seminar on Multi-modal AI for Medicine (IN2107, IN45072). #AIMnews

110

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 16/12/2025

🎄 Lab Christmas Party! 🎄 It's always great to use the holiday season as an opportunity to (re-)connect! We had a blast at our Christmas party with lots of laughs, pizza, and Glühwein 🍷 Wishing everyone a good end to this year and happiest of holidays to those who celebrate! ✨ #AIMsocial

030

Reposted by TUM AI in Medicine Lab

Florian Hölzl @hlzl.eurosky.social · 10/12/2025

Great being at #NeurIPS last week! Thankful for the good times in the sun and the people I met. If you’ve ever wondered whether model performance can be inferred directly from your training alone, check out our work! 1/2

141

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 02/12/2025

It’s this time of the year again! #NeurIPS2025 If you are in San Diego, make sure to check out 2 works from our lab this week 📝📝 #AIMresearch 1/3

133

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 27/11/2025

Huge congratulations to Paul Hager and Dr. med. Friederike Jungmann for winning the MDSI Best Paper Award in the Societal Impact category! 🏆🎊 See more details in the thread below! Photo's copyright: Andreas Heddergott/TUM #AIMResearch #AIMNews #MDSI #BestPaperAward

140

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 03/11/2025

We celebrated the 5th anniversary of our research chair at @tum.de! 💙🥂 It's been an incredible journey of research and collaboration. Thank you to everyone who has made this possible. We are very much looking forward to the next years to come! #AIMAnniversary #AIMNews

052

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 29/10/2025

Quick look back at an insightful day yesterday at the Bavarian Conference on AI in Medicine, where @paulhager.bsky.social , @luciehuang.bsky.social, Alina Dima, and Vasiliki Sideri-Lampretsa were representing us. Team, thank you for being such good ambassadors! 👏 #AIMNews #AIinMedicine

021

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 27/10/2025

Is it safe to use LLMs in the clinic today? This is the central question that @paulhager.bsky.social and Friederike Jungmann tackled in their 2024 study published in #NatureMedicine, which is our 5th-year anniversary's highlight paper this week. #AIMAnniversary #AIMResearch #LLM #Benchmark

100

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 23/10/2025

Last week we had the pleaure of hosting Paul Kaftan (Institute of Medical System Biology, @uniulm.bsky.social) and Steven Jia (Institut de Neurosciences de la Timone, @univ-amu.fr) for talks on the incredible potential of Implicit Neural Representations (INRs) in medical imaging. #AIMnews #INRs

130

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 17/10/2025

Huge and well-deserved congratulations to Dmitrii Usynin on successfully completing his doctoral journey in our lab! 🎊 Dima’s research has resulted in significant contributions to field of trustworthy artificial intelligence in particular for collaborative biomedical image analysis.

130

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 09/10/2025

Throwback to 2022! 🕰️ Continuing with our 5th-year anniversary series, we're now revisiting a question we faced at that time: How fast can we reliably see the beating heart in MRI without sacrificing image quality? #AIMAnniversary #AIMResearch #NeuralImplicitRepresentation #CardiacMRI

100

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 06/10/2025

We are thrilled to announce that Dr. Sevgi Gokce Kafali has been awarded the prestigious Alexander von Humboldt Postdoctoral Fellowship! 🎉 During her fellowship, she will focus on the "Assessment of Cardiac Health through Opportunistic Screening using MRI." Our warmest congratulations! #AIMNews

000

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 02/10/2025

As our lab's 5th anniversary approaches 🥳, we're kicking off a month-long series of posts highlighting standout papers from our past five years at @tum.de. Join us as we revisit the incredible work that has shaped our journey! #AIMAnniversary #AIMResearch

130

Reposted by TUM AI in Medicine Lab

Julia Schnabel @ja-schnabel.bsky.social · 01/10/2025

Submit to our @euripsconf.bsky.social workshop : MedEurIPS - Medical Imaging meets EurIPS sites.google.com/view/medeurips @aasaferagen.bsky.social @benglocker.bsky.social @bernhardkainz.bsky.social @danielrueckert.bsky.social Ender Konukoglu @stsaftaris.bsky.social

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 19/09/2025

The annual conference of the @miccaisociety.bsky.social is in 3 days! 🗓️ Our lab is excited to present 8 papers exploring advancements in AI for medical imaging 🧠🩻 Stop by our posters & presentations to learn more! ➡️ conferences.miccai.org/2025/ Overview of our papers in the thread below 🧵⤵️

conferences.miccai.org

MICCAI 2025 - 28. International Conference On Medical Image Computing & Computer Assisted Intervention

MICCAI 2025, the 28th International Conference on Medical Image Computing and Computer Assisted Intervention, will be held from September 23rd to 27th 2025 in Daejeon/Republic of Korea.

161

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 10/09/2025

🩻 Are you a student at @tum.de and interested in AI and medicine? Make sure to check out our courses for the upcoming winter semester! Practical: Applied Deep Learning in Medicine (IN2106, IN4314) Lecture: Artificial Intelligence in Medicine (IN2403) ...

120

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 26/08/2025

🚀 Towards cardiac MRI foundation models: Comprehensive visual-tabular representations for whole-heart assessment and beyond. We introduce ViTa, a multi-view, multi-modal, multi-task model for cardiac MRI. 🧵 1/3

121

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 20/08/2025

🔃 We're moving beyond traditional benchmarking for LLMs. 🚩 Meet Dynamic, Automatic & Systematic (DAS) Red-Teaming from Jiazhen Pan and Bailiang Jian together with great collaborators! #AIMresearch 🧵 1/4

Outline of the Dynamic, Automatic & Systematic (DAS) Red‑Teaming method.

131

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 18/08/2025

🤔 Ever wondered what a ‘reference brain’ would look like for your own demographic group? @starcksophie.bsky.social and @siderilampretsa.bsky.social, together with some amazing colleagues, have developed a novel diffusion-based approach for this. But how does this work? #AIMresearch 1/3

131

Reposted by TUM AI in Medicine Lab

Julian Suk @sukjulian.bsky.social · 04/08/2025

I am happy to announce that I have started a postdoc position at @aim-lab.bsky.social and @munichcenterml.bsky.social, Technical University of Munich under the supervision of @danielrueckert.bsky.social. Looking forward to this new chapter 🥨🎓

051

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 01/08/2025

Last week, we had the pleasure of hosting Prof. @clemensdlaska.bsky.social and part of his research team: Nadja Gruber (PhD), Angus Nicolson (PhD), and Riccardo Lunelli. Prof. Dlaska gave a talk on "Digital Cardiology: From Benchmarks to Bedside". #AIMnews #DigitalCardiology

120

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 29/07/2025

Huge and well-deserved congratulations to Philip Müller on successfully completing his doctoral journey in our lab! 🎊 Philip’s dedication and innovative research have led to significant contributions in localized image-text learning for chest X-rays. scholar.google.com/citations?us...

120

Reposted by TUM AI in Medicine Lab

Technische Universität München @tum.de · 25/07/2025

As part of Daniel Rückert's team, our #PhDcandidate Vasiliki Sideri-Lampretsa uses #AI and high-performance #computing to analyze MRI & CT scans, tracking lung movement and detecting #diseasepatterns: go.tum.de/395577 #clinicalpractice 📷A.Eckert

A woman stands smiling with arms crossed against a building facade, with a large sculpture resembling a DNA helix visible in the background.

A woman sits at a desk working on a laptop while medical imaging data, such as MRI brain scans, is displayed on a large monitor.

041

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 23/07/2025

As of yesterday, our research team gathered for a 4-day work retreat in Obertraun, Austria. It was a fantastic opportunity to connect through meaningful keynotes, strategy sessions, team-building activities, sports, and even a (legendary!) pub quiz. #WorkRetreat #StrategySessions #SocialActivities

140

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 17/07/2025

We recently had the privilege of hosting George Kaissis as a guest speaker in our Multimodal Deep Learning lecture at @tum.de. George, currently at Google DeepMind, guided us from the first principles to the state of the art in AI video generation. #Teaching #MultimodalDeepLearning

240

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 09/07/2025

Our PhD students, Alina D. and @luciehuang.bsky.social, are participating in #ICVSS2025. During the poster session, Alina presented her research on parametric shape models of 3D vessels learned from segmentations using differentiable voxelization (arxiv.org/pdf/2507.02576).

050

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 04/07/2025

One year ago, @martinmenten.bsky.social established the AI for Vision group at our chair. 🎉 Supported by funding from the Emmy Noether Programme of the @dfg.de and the @munichcenterml.bsky.social , the group has since grown to six researchers! 1/2

170

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 27/06/2025

Groundbreaking ceremony complete for the new Zentrum für Digitale Medizin und Gesundheit (#ZDMG)! This interdisciplinary research center is a major step in further strengthening digital medicine research & developing future clinical approaches @tum.de. 🎉

120

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 25/06/2025

Pizza + Sunshine + Good Company = Optimal Setting for our lab‘s summer fest today! ☀️🍕 #VitaminD

062

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 12/06/2025

Responsible medical AI demands patient-level privacy. Our recent #TPDP '25 paper extends Differential Privacy #DP beyond individual data points to protect entire patient profiles. 🧠 📄 Read the paper: tinyurl.com/k9fz456a 🎥 Watch the video: tinyurl.com/2jx528m5

tpdp.journalprivacyconfidentiality.org

031

Reposted by TUM AI in Medicine Lab

Paul Hager @paulhager.bsky.social · 11/06/2025

Going to be at CVPR the next couple of days presenting our paper „A Tale of Two Classes: Adapting Supervised Contrastive Learning to Binary Imbalanced Datasets“. arxiv.org/abs/2503.17024 Always happy to meet anyone working on representation learning or tabular DL and medical data

arxiv.org

A Tale of Two Classes: Adapting Supervised Contrastive Learning to Binary Imbalanced Datasets

Supervised contrastive learning (SupCon) has proven to be a powerful alternative to the standard cross-entropy loss for classification of multi-class balanced datasets. However, it struggles to learn ...

073

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 08/06/2025

Huge congratulations to Jiazhen on successfully completing his doctoral journey in our lab! 🎉 His work and dedication have resulted in great contributions to cardiac MRI reconstruction for faster and higher-quality imaging. Good luck on your future endeavors, JZ!

131

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 28/05/2025

If you're at #IPMI in Greece 🇬🇷 right now, make sure to check out 2 works of our group tomorrow, Thursday! Alex and Laurin are discussing pitfalls in their research on topology-aware image segmentation... (1/2) arxiv.org/abs/2412.14619

arxiv.org

Pitfalls of topology-aware image segmentation

Topological correctness, i.e., the preservation of structural integrity and specific characteristics of shape, is a fundamental requirement for medical imaging tasks, such as neuron or vessel segmenta...

142

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 22/05/2025

We are incredibly proud of Prof @danielrueckert.bsky.social for being elected as Fellow of the Royal Society! Well deserved and a testament to your dedication to research 🎉

182

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 18/05/2025

#ISMRM 2025 took place this week, and here is a brief recap of our PhD students' presentations. 1/n #CardiacMRI #ImplicitNeuralRepresentations #RepresentationLearning

120

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 15/05/2025

We had the pleasure of hosting Thomas Hadler from Charité - Universitätsmedizin Berlin and Kirsten Maas from @tue.nl in our lab this week. Both gave insightful talks on their research.

120

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 09/05/2025

Huge congratulations to Felix Meissen on successfully completing his doctoral journey in our lab! 🎉 His hard work and dedication have paid off and represented contributions to medical pathology detection using weak supervision in Machine Learning. We wish him all the best in his future projects.

0121

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 08/05/2025

Are you working with datasets that follow the Brain Imaging Data Structure (BIDS) naming convention? If so, you may want to try TPTBox, a toolbox developed by our PhD students Hendrik Möller and Robert Graf. It offers a wide variety of tools for working with NIfTI files. github.com/Hendrik-code...

github.com

GitHub - Hendrik-code/TPTBox: Torso Processing ToolBox

Torso Processing ToolBox. Contribute to Hendrik-code/TPTBox development by creating an account on GitHub.

000

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 28/04/2025

#ICLR 2025 has concluded, and here is a brief recap of our PhD students' presentations. Alexander Berger and Laurin Lux presented their work on topology-preserving image segmentation, while Johannes Kaiser discussed his work on data attribution. For details, check out our previous two posts.

020

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 20/04/2025

Data attribution is crucial for debugging models and detecting low quality data (spotting mislabeled samples, bias etc.). But many methods aren't mathematically sound and don’t scale. But how could we improve this for large models? 1/n

141

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 16/04/2025

Topological accuracy in image segmentation is crucial for many tasks - such as blood flow or functional brain analysis. Yet, most models disregard this structural integrity during training, leading to potential inaccuracies in critical analyses. But how could we address this issue? 1/n

152

TUM AI in Medicine Lab @tum-aim-lab.bsky.social · 13/04/2025

Looking for MRI spine segmentation? Try SPINEPS, a state-of-the-art framework for automatic spine segmentation by our PhD student, Hendrik Möller. It annotates 14 structures both semantically and instance-wise, simplifying downstream analysis. github.com/Hendrik-code...

github.com

GitHub - Hendrik-code/spineps: This is a segmentation pipeline to automatically, and robustly, segment the whole spine in T2w sagittal images.

This is a segmentation pipeline to automatically, and robustly, segment the whole spine in T2w sagittal images. - Hendrik-code/spineps

010

Reposted by TUM AI in Medicine Lab

Paul Hager @paulhager.bsky.social · 01/04/2025

Excited to share: “A Tale of Two Classes: Adapting Supervised Contrastive Learning to Binary Imbalanced Datasets” has been accepted to #CVPR2025! 🎉 Paper: lnkd.in/esKRqF5p Code: lnkd.in/eZFvDA5Q (Thread incoming 👇)

1125