Computational Medical Imaging |

Stain Normalization for Histology Images

Fri, 30 Oct 2026 00:00:00 +0000

Placeholder — the interactive demo is in preparation.

Colour variation between labs and slide scanners is one of the main reasons histology models fail on new sites. This demo will let you upload an H&E image tile and transfer it to a reference stain style, comparing a learned approach ( ) with classical stain-normalisation methods. See also the .

MRIQT — Image-Quality Transfer for Ultra-Low-Field MRI

Fri, 30 Oct 2026 00:00:00 +0000

Slide to compare the portable ultra-low-field scan (uLF, right) with the MRIQT enhancement (left). Use the buttons to switch between two patients.

MRIQT uLF (0.064 T)

Neonatal brain MRI from the University Hospital Bonn cohort. Ethical approval: Ethics Committee of the University of Bonn (Nr. 167/22).

· · · Accepted at , London

Why it matters

Infants change fast in the first weeks of life: their brains differ from adult brains in both anatomy and tissue contrast. In neonatal intensive care units, many infants have neurological conditions that need repeated MRI, yet standard high-field MRI means disconnecting them from life support and moving them to a scanner room — slow, often requiring sedation and exposure to loud noise.

Portable ultra-low-field MRI (0.064 T) brings the scanner to the bedside, but with low signal-to-noise and reduced diagnostic quality. MRIQT closes that gap: it performs 3D image-quality transfer (uLF → high-field) with a conditional diffusion model guided by physics-aware k-space simulation and an anatomy-preserving 3D perceptual objective. The cohort comprises neonates with an average age of two months and nine clinically relevant neurological pathologies, often occurring together.

How it works

Physics-aware degradation: realistic ultra-low-field scans are simulated from high-field ones in k-space.
3D conditional diffusion: the model enhances the scan while preserving anatomy.
Guidance and an SNR-weighted 3D perceptual loss reduce artefacts and hallucinations.

Training — MRIQT learns from physics-based uLF simulations to reconstruct anatomically faithful images from noisy inputs:

Inference — given a real uLF scan, MRIQT iteratively denoises a noisy initialisation while conditioning on the uLF measurement:

Results

Compared with competing methods on PSNR, LPIPS, MAE, MS-SSIM, SSIM and Pearson correlation:

Echocardiography Analysis

Fri, 30 Oct 2026 00:00:00 +0000

Placeholder — the interactive demo is in preparation. A description, the model and the accompanying paper will be added here.

Dose Generation for Radiation Therapy

Fri, 30 Oct 2026 00:00:00 +0000

Placeholder — the interactive demo is in preparation.

Online adaptive radiotherapy re-plans treatment while the patient is on the table, which requires dose distributions within seconds. This demo will generate a dose distribution from the planning image and contoured structures, developed within the DFG project with LMU and TU Munich.

AIR — Winter 2026/27

Thu, 15 Oct 2026 00:00:00 +0000

Seminar (2 SWS): Artificial Intelligence in Radiology: Applications and Research (Artifizielle Intelligenz in der Radiologie: Anwendungen und Forschung)

Lecturers:

Tutors: to be announced

Time: November 15 – Jan 31 (every Friday, ~~11:30 AM - 1:00 PM~~ 16:00 - 17:30)

Format: Online (Zoom Meeting) with potential face-to-face meetings, if participants interested! Link was provided via email.

Language: English

eCampus: to eCampus

Description

The Computational Imaging Research (Albarqouni Lab) at the Clinic of Diagnostic and Interventional Radiology at the University Hospital Bonn is pleased to present a seminar on Artificial Intelligence in Radiology: Applications and Research. This cutting-edge seminar, spanning two semester hours per week (2 SWS), is specially curated for medical students to explore the transformative role of deep learning in medical imaging, with a focus on tasks such as classification, detection, segmentation, reconstruction, tracking, and disease progression.

Course Objectives

Upon completing this seminar, students will be able to:

Grasp deep learning concepts and their specific applications in medical imaging.
Analyze advanced image analysis tasks, including classification, detection, segmentation, and disease progression modeling.
Evaluate ethical and regulatory considerations in the deployment of AI within healthcare.
Discuss the future potential of AI-driven solutions in personalized medicine and patient care.

Target Audience

This seminar is designed for medical students at the Faculty of Medicine, University Hospital Bonn, who are keen on leveraging AI technologies to enhance their medical practice, specifically within diagnostic imaging.

Prerequisites and Desired Knowledge

Prerequisites: Familiarity with fundamental medical imaging techniques and an interest in AI applications.
Desired: Understanding of medical diagnostics and clinical data interpretation.

Seminar Topics and Schedule (Tentative)

Date	Topic	Description	Materials
TBA	Deep Learning Foundations	Introduction to deep learning concepts specific to medical imaging.	Meskó, Bertalan, and Marton Görög. “A short guide for medical professionals in the era of artificial intelligence.” NPJ digital medicine 3.1 (2020): 126.
TBA	Detection and Classification in Medical Imaging	Exploring detection and classification techniques in diagnosing diseases from medical images.	Litjens, Geert, et al. “A survey on deep learning in medical image analysis.” Medical image analysis 42 (2017): 60-88. , Chen, Xuxin, et al. “Recent advances and clinical applications of deep learning in medical image analysis.” Medical image analysis 79 (2022): 102444. ,
TBA	Segmentation	Deep learning methods for segmenting key structures in medical images.	Tajbakhsh, Nima, et al. “Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation.” Medical image analysis 63 (2020):101693.
TBA	Causality and Bias	Causality and Bias in deep learning with medical imaging.	Castro, Daniel C., Ian Walker, and Ben Glocker. “Causality matters in medical imaging.” Nature Communications 11.1 (2020): 3673. , Jones, Charles, et al. “A causal perspective on dataset bias in machine learning for medical imaging.” Nature Machine Intelligence 6.2 (2024): 138-146
TBA	~~Uncertainity Quantification~~	Moved to the next year due to holiday season
TBA	Q&A Session	Students should prepare questions for their assigned papers and discuss them with the tutors
TBA	Uncertainity Quantification	Quanitifying Uncertainty in Deep Learning Models	Huang, Ling, et al. “A review of uncertainty quantification in medical image analysis: probabilistic and non-probabilistic methods.” Medical Image Analysis (2024): 103223.
TBA	Student Paper Presentations (1)	Student presentations.
TBA	Student Paper Presentations (2) and Course Wrap-Up	Wrap-up, feedback, and discussion on future AI applications in medical practice.

Learning Format

One of the seminar’s unique teaching concepts is “Teach It to Yourself.” Participants (limited to 20 partisipants) will present selected chapter and papers, engage in discussions, and critically analyze case studies on deep learning applications in medical imaging.

Registration (closed)

Interested in this seminar, please register via email to and

References

[1] Litjens, Geert, et al. “A survey on deep learning in medical image analysis.” Medical image analysis 42 (2017): 60-88.
[2] Shen, Dinggang, Guorong Wu, and Heung-Il Suk. “Deep learning in medical image analysis.” Annual review of biomedical engineering 19.1 (2017): 221-248.
[3] Chen, Xuxin, et al. “Recent advances and clinical applications of deep learning in medical image analysis.” Medical image analysis 79 (2022): 102444.
[4] Tajbakhsh, Nima, et al. “Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation.” Medical image analysis 63 (2020):101693.
[5] Wang, Ge, Jong Chul Ye, and Bruno De Man. “Deep learning for tomographic image reconstruction.” Nature machine intelligence 2.12 (2020): 737-748.
[6] Ben Yedder, Hanene, Ben Cardoen, and Ghassan Hamarneh. “Deep learning for biomedical image reconstruction: A survey.” Artificial intelligence review 54.1 (2021): 215-251.
[7] Huang, Ling, et al. “A review of uncertainty quantification in medical image analysis: probabilistic and non-probabilistic methods.” Medical Image Analysis (2024): 103223.
[8] Castro, Daniel C., Ian Walker, and Ben Glocker. “Causality matters in medical imaging.” Nature Communications 11.1 (2020): 3673.
[9] Jones, Charles, et al. “A causal perspective on dataset bias in machine learning for medical imaging.” Nature Machine Intelligence 6.2 (2024): 138-146.
[10] Thirunavukarasu, Arun James, et al. “Large language models in medicine.” Nature medicine 29.8 (2023): 1930-1940.
[11] Zhang, Jingyi, et al. “Vision-language models for vision tasks: A survey.” IEEE Transactions on Pattern Analysis and Machine Intelligence (2024).

CoMI-L — Winter 2026/27

Thu, 15 Oct 2026 00:00:00 +0000

Lab: Computational Medical Imaging (CoMI)

Lecturers:
Tutors:
Module: MA-INF 2230
Credits: 4 SWS, 9 CP (270h workload)
Programme: M.Sc. Computer Science
Mode: Optional (2. or 3. Semester)
Language: English
Format: Lab (Praktikum), groups of max. 3 students
Frequency: Every year

Description

The Computational Medical Imaging (CoMI) Lab is a project-based course in which students carry out a semester-long research project in computational medical imaging. Working in small groups, students identify a well-defined problem, collect or are provided a suitable dataset, implement baseline and novel approaches, and evaluate their results both quantitatively and qualitatively.

Possible project topics (chosen or proposed by the instructors) include:

Anatomical Structure Segmentation (e.g., multi-atlas, CNN-based)
Deformable Registration (e.g., classic demons, learning-based registration)
Lesion/Anomaly Detection (e.g., out-of-distribution detection in X-ray)
Radiomics & Quantitative Biomarker Extraction (e.g., texture features, deep radiomics)
Uncertainty Estimation in Deep Learning (e.g., MC-Dropout, Bayesian networks)
Causality & Bias Analysis in Medical Datasets (e.g., domain shift, fairness)

Learning Goals

Technical Skills

Acquire a deep understanding of a selected problem in computational medical imaging: from problem identification to data processing, algorithm design, implementation, and evaluation.
Implement, test, and validate state-of-the-art image-analysis algorithms (e.g., segmentation, registration, reconstruction, classification) on real medical data.
Develop proficiency in using libraries and toolkits commonly used in medical imaging (e.g., ITK, SimpleITK, PyTorch for deep learning, nibabel/dipy).
Learn to document software systematically (API documentation, code comments, unit tests) and to generate reproducible experiments (version control, containerization).

Soft Skills

Collaborate effectively in small groups over the entire semester, dividing tasks (e.g., data curation, algorithmic implementation, evaluation).
Prepare clear, concise technical documentation and design rationales.
Present project progress and final results in oral form (midterm demo, final demo).
Critically assess one’s own results and position them within the state-of-the-art in computational medical imaging.

🗓 Project (Winter 2026/27)

Who Does the Model See? Bias of Ultrasound Foundation Models across Countries and Income Settings in Fetal Ultrasound

Ultrasound is the imaging modality of antenatal care worldwide, and in low- and middle-income countries scans are increasingly done by non-specialists on low-cost devices. Ultrasound foundation models promise label-efficient AI support exactly there — but they are pretrained almost entirely on data from high-income hospitals. Students build a harmonised benchmark from 12 openly licensed fetal-ultrasound releases (13 countries, four World Bank income groups) and test whether frozen foundation models work equally well for fetuses scanned in Malawi, Uganda, Sierra Leone, Ghana, Egypt, Algeria, Brazil and China as in Spain, the Netherlands, Romania and the UK.

Bias metrics: worst-group performance, between-group gaps and calibration, with patient-level confidence intervals.
Clinical highlight: a measurement study — do errors in the cephalic index, abdominal circumference and cardiothoracic ratio, and so the share of fetuses flagged as abnormal, differ by site, device and income group?
Models: FetalCLIP, USFM, SonoNet-64 (USF-MAE secondary).
Team: 3 students; all share data harmonisation and the final synthesis, each leads one experimental work package. Students implement their own toolkit (fusbias) against a provided specification and acceptance tests.

WP	Topic	Lead
1	Data harmonisation, de-duplication, patient-level splits and quality control	all
2	Zero-shot and training-free standard-plane recognition	S1
3	Frozen-feature probing and bias audit	S2
4	Clinical measurements: cephalic index, abdominal circumference, cardiothoracic ratio	S3
5	Representation analysis, synthesis and report	all

Learning Format

Students work in groups of up to 3 on a semester-long project. Each group will:

Identify a well-defined problem and collect or be provided a suitable dataset (publicly available or hospital data under ethics approval).
Perform a literature review to select baseline methods.
Implement baseline and at least one novel or improved approach.
Evaluate quantitatively (e.g., Dice, Hausdorff, MSE, classification accuracy) and qualitatively (expert review where possible).
Document the software (README, API docs) and provide a reproducible environment (Docker/Singularity).
Present midterm progress (oral demonstration + short written summary) and final results (oral + poster or full report).

Assessment

Graded Exam:
Oral presentation (final demo + Q&A, 20 minutes per group). Written report (12–16 pages, including background, methods, results, discussion).

Ungraded Coursework (required for admission to the exam):
Project work; attendance in course sessions in accordance with the exam regulations of 2023, § 12(6).

Recommended Prerequisites

At least one of the following:

MA-INF 2222 – Visual Data Analysis
MA-INF 2312 – Image Acquisition and Analysis in Neuroscience
MA-INF 2220 – Lab Visualization and Medical Image Analysis (or equivalent)

A solid background in Python programming (and/or C++), linear algebra, calculus, probability theory, and numerical algorithms is required. Familiarity with deep learning frameworks is highly recommended.

References and Resources

Relevant Journals:

Nature Machine Intelligence
Medical Image Analysis (MedIA)
IEEE Transactions on Medical Imaging (TMI)
Radiology: AI, European Radiology, IEEE Journal of Biomedical and Health Informatics

Relevant Conferences:

MICCAI (Medical Image Computing & Computer-Assisted Intervention)
CVPR (IEEE/CVF Computer Vision and Pattern Recognition)
MIDL (Medical Imaging with Deep Learning)
ISBI (International Symposium on Biomedical Imaging)
RSNA (Radiological Society of North America Annual Meeting)

Registration

Registration takes place via eCampus or by contacting .

Medical Pitch Contest Bonn

Wed, 14 Oct 2026 00:00:00 +0000

attended MIB Future Panel 2026 where he also pitched , along with , at the Medical Pitch Contest and our plan to commercialze the project. The MIB Future Panel is UKBonn’s annual networking event hosted by the Medical Imaging Center and UKBonn’s Transfer Team in Bonn. The goal of this event is to merge medical imaging and artificial intelligence (AI) in clinical practice and research with a panel discussion with specilists. The bulk of the event features a medical pitch contest featuring teams from the medical faculty at the University Hospital Bonn, where venture capital investors serve as a jury, evaluating the competing teams projects for a competing prize of €60,000. The event offers an opportunity for teams to network with investors and accelerator programs to move from the research phase into commercialization. Our work, MRIQT-Neo is a software that works on image quality transfer for MRI images, from the portable ultra-low-field MRI into high-quality, 3T- like images. The networking and feedback event provided a valuable feedback for us on how to

K-MaT: Knowledge-Anchored Manifold Transport for Cross-Modal Prompt Learning in Medical Imaging

Sun, 27 Sep 2026 00:00:00 +0000

Our Lab at MICCAI 2026 in Strasbourg

Sun, 27 Sep 2026 00:00:00 +0000

Our lab was at MICCAI 2026 in Strasbourg, France! This year, we contributed across research, workshops, and the broader MICCAI Society. It was great to see these efforts come together, from developing new methods to creating spaces for discussion, collaboration, and exchange.

Papers at MICCAI

Our lab had two papers accepted at the main conference:

— Adea Nesturi, David Dueñas Gaviria, Jiajun Zeng, and Shadi Albarqouni

Adea Nesturi and David Dueñas Gaviria presented the work as poster W-AM-192. The paper introduces PromptGate, a vision-language gating framework that filters out-of-distribution samples before active learning. This helps reduce unnecessary annotation and improves the quality of selected medical images while preserving patient privacy across institutions.
— Jiajun Zeng and Shadi Albarqouni

Jiajun Zeng presented K-MaT as poster W-AM-193. The work studies high-to-low generalization in medical imaging, transferring knowledge from high-end to low-end modalities without access to low-end labels. These posters attracted considerable interest and led to many engaging discussions with researchers at the conference.

Co-organized workshops

We also co-organized two workshops:

(September 27), with Yasmeen George, Dwarikanath Mahapatra, Imran Razzak, among others.
(October 1), with Holger R. Roth, Spyridon Bakas, Xiaoxiao Li, Yasmeen George, Meirui Jiang, and others.

Workshop Satellite Event

In addition, Prof. Shadi Albarqouni served as one of the Workshop Chairs of MICCAI 2026, overseeing the workshop satellite events that brought together researchers and organizers from across the community.

Overall, it was a great week for the group: reconnecting with familiar faces, meeting new people, and having many interesting conversations throughout the conference. We look forward to the next one!

PromptGate: Client-Adaptive Vision–Language Gating for Open-Set Federated Active Learning

Sun, 27 Sep 2026 00:00:00 +0000

Two Awards to Attend MICCAI 2026 in Strasbourg

Sun, 27 Sep 2026 00:00:00 +0000

We are delighted to share that two members of our group have received awards supporting their participation in , the 29th International Conference on Medical Image Computing and Computer Assisted Intervention, taking place from September 27 to October 1, 2026, in Strasbourg, France.

David has been selected to receive the 2026 MICCAI Society Registration Grant, awarded by the MICCAI Society through its program. The grant is awarded to students and early-career scientists who are authors of an accepted MICCAI 2026 paper and covers the in-person main conference registration, both days of satellite event registrations, and one year of MICCAI Society membership. The MICCAI Society awards approximately 25 Registration Grants per year across the whole community.

Adea has been awarded funding under the University of Bonn’s program for doctoral researchers, administered by the Bonn Graduate Center. The award covers travel, accommodation, and conference participation fees for doctoral students actively presenting their work at an international conference.

Both awards will be used to attend MICCAI 2026 in Strasbourg, where they will present our accepted paper “PromptGate: Client-Adaptive Vision–Language Gating for Open-Set Federated Active Learning”. MICCAI 2026 received a record number of submissions this year: .

We thank the MICCAI Society and the University of Bonn for their support, and we look forward to presenting our work and meeting the community in Strasbourg.

Entwicklung einer künstlichen Intelligenz zur ultraschallgestützten Detektion der Riesenzellarteriitis

Wed, 09 Sep 2026 00:00:00 +0000

Invited Talk at Karman Conference on Innovations in non-invasive and computational imaging

Mon, 07 Sep 2026 00:00:00 +0000

I had the pleasure of delivering an invited talk on Foundation Models in Medicine at the Karman Conference on Innovations in non-invasive and computational imaging, . Interested in hearing about our recent work? Come to in Strasbourg . My team members will present their accepted papers on Active Learning and Multi-modal data!

When multi-institutional dataset aggregation masks shortcut-like behavior in Foundation Models for medical imaging

Tue, 01 Sep 2026 00:00:00 +0000

MISTAI: Medical Intelligence & Skills Training in AI

Wed, 01 Jul 2026 00:00:00 +0000

The war in Ukraine has severely disrupted medical research and teaching. Digital pathology and AI offer a way to keep both going without depending on physical infrastructure. MISTAI is a bilateral cooperation between the University Hospital Bonn and Dnipro State Medical University (DSMU), funded by the Ministry of Culture and Science of North Rhine-Westphalia (MKW NRW).

Goals

Shared data infrastructure: digitise and validate at least 100 H&E and IHC whole-slide images from a DSMU cohort of muscle-invasive bladder cancer (MIBC), on a joint secure platform.
AI diagnostic pipeline: fine-tune state-of-the-art pathology foundation models on the combined Bonn–Dnipro data, including spatial analysis of tumour-infiltrating lymphocytes from routine H&E slides, inspired by the UC-TIL approach.
Clinical tool: a QuPath plugin that lets pathologists run the trained model directly in their existing workflow.
Capacity building: weekly remote collaboration, a one-week exchange visit to Bonn, and a Computational Pathology teaching module for the DSMU curriculum.
Follow-on work: a joint DFG/BMBF proposal outline and a MICCAI challenge concept.

Partners

Bonn contributes high-resolution slide scanners, AI infrastructure and expertise in computational medical imaging (Prof. Shadi Albarqouni) and translational lung and tumour pathology (Prof. Verena Tischler). DSMU contributes a clinically distinct bladder-cancer cohort and long-standing pathology expertise (Assoc. Prof. Nina Bondarenko, Department of Pathological Anatomy).

PromptGate

Sat, 20 Jun 2026 00:00:00 +0000

Overview

PromptGate tackles open-set federated active learning, where each client holds a private unlabeled pool contaminated with out-of-distribution (OOD) samples and artifacts. The goal is to spend a limited annotation budget on in-distribution (ID) candidates while keeping client data private.

At the core is a VLM gate built on a frozen BiomedCLIP backbone (image + text encoders). Cosine similarity between image features and prompt embeddings yields VLM-gated pseudo-labels that separate the unlabeled pool into ID candidates and OOD / artifacts, so the active-learning oracle is only queried on likely-ID samples.

Gating is driven by client-adaptive learnable prompts, which decompose into:

Global tokens — shared across clients via FedAvg.
Local tokens — kept private and never leave the site.
Mixed tokens — combine global and local context (with a Class-Specific-Context / CSC variant), replacing the static text-prompt baseline.

A few shots of ID/OOD labeled data tune these prompts via CoOp. The downstream task model runs each active-learning round (inference → training), and its parameters are shared via FedAvg, while a clinician oracle labels the queried ID candidates. PromptGate is agnostic to the acquisition rule and is evaluated with Random, Entropy, FEAL, PAL, LfOSA, and OpenPath.

Using the code

The repository README covers environment setup, install the remaining requirements, datasets & attribution, out-of-distribution (ood) setup, usage. See for step-by-step instructions.

BONFOR Symposium & Promotionsbörse in Bonn

Tue, 16 Jun 2026 00:00:00 +0000

David represented our group at the BONFOR Symposium 2026 in Bonn, taking part in the Promotionsbörse (doctoral thesis fair) organized by the Medical Student Council alongside the symposium at the BMZ I. We have presented the work of the Computational Medical Imaging Research group (Albarqouni AG) at the Clinic for Diagnostic and Interventional Radiology and shared what doing a doctoral thesis with us looks like. It is a great opportunity to connect with the medical students looking for a thesis topic, walk them through our research directions in medical imaging and machine learning, and address their questions about getting started in research.

ArtFusion — Official codebase for the ArteryFusion model

Tue, 19 May 2026 00:00:00 +0000

Using the code

The repository README covers 1. training. See for step-by-step instructions.

OMI Plenary Meeting in Erlangen

Thu, 23 Apr 2026 00:00:00 +0000

David and Malek attended the OMI Plenary Meeting in Erlangen 2026, where partners from across the consortium gathered to share progress, discuss issues and fixes, and plan the next steps. The meeting hosted at the Digital Health Innovation Platform in Erlangen, featured an outstanding talk by Prof. Bernhard Kainz about the future of AI in healthcare and how OMI can be a great support, not only for researchers, but also a great potential for physcians. The breakout sessions featured intensive and insightful technical discussions, while the individual work package updates demonstrated that all partner sites are progressing according to schedule. A few key takeaways from the two days were: 1) Meeting everyone in person, along with side conversations and open group discussions, led to a deeper understanding of OMI and deepened understanding among partners, 2) key challenges include data and server hosting security, the responsibility and cost of hosting the service and safeguarding the data, and asynchornous updating of the individual components easily break compatibility with other components, 3) despite the regulatory and technical challenges, OMI shows strong potential to enable open yet secure data and knowledge sharing across partners, provided that the necessary security agreements and guarantees are established, 4) the commitment of everyone involved has driven the project to reach, to a large extent, the expected maturity level according to the timeline, allowing actual services to be tested.

AIR — Summer 2026

Mon, 13 Apr 2026 00:00:00 +0000

Seminar (2 SWS): Artificial Intelligence in Radiology: Applications and Research (Artifizielle Intelligenz in der Radiologie: Anwendungen und Forschung)

Lecturers:

Tutors:

Module: 401040415

Credits: 2 SWS

Programme: Medical Students at Uni Bonn

Language: English

Format: Seminar (2 SWS)

Time: Fridays, 09:00 – 11:00

Location: Online

Description

Course Objectives

Upon completing this seminar, students will be able to:

Grasp deep learning concepts and their specific applications in medical imaging.
Analyze advanced image analysis tasks, including classification, detection, segmentation, and disease progression modeling.
Evaluate ethical and regulatory considerations in the deployment of AI within healthcare.
Discuss the future potential of AI-driven solutions in personalized medicine and patient care.

Target Audience

Prerequisites and Desired Knowledge

Prerequisites: Familiarity with fundamental medical imaging techniques and an interest in AI applications.
Desired: Understanding of medical diagnostics and clinical data interpretation.

Seminar Topics and Schedule (Tentative)

Date	Topic	Description	Materials
May 15	Deep Learning Foundations	Introduction to deep learning concepts specific to medical imaging.	Meskó, Bertalan, and Marton Görög. “A short guide for medical professionals in the era of artificial intelligence.” NPJ digital medicine 3.1 (2020): 126.
June 12	Detection and Classification in Medical Imaging	Exploring detection and classification techniques in diagnosing diseases from medical images.	Litjens, Geert, et al. “A survey on deep learning in medical image analysis.” Medical image analysis 42 (2017): 60-88. , Chen, Xuxin, et al. “Recent advances and clinical applications of deep learning in medical image analysis.” Medical image analysis 79 (2022): 102444. ,
July 10	Segmentation	Deep learning methods for segmenting key structures in medical images.	Tajbakhsh, Nima, et al. “Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation.” Medical image analysis 63 (2020):101693.
	Causality and Bias	Causality and Bias in deep learning with medical imaging.	Castro, Daniel C., Ian Walker, and Ben Glocker. “Causality matters in medical imaging.” Nature Communications 11.1 (2020): 3673. , Jones, Charles, et al. “A causal perspective on dataset bias in machine learning for medical imaging.” Nature Machine Intelligence 6.2 (2024): 138-146
	Uncertainity Quantification	Quanitifying Uncertainty in Deep Learning Models	Huang, Ling, et al. “A review of uncertainty quantification in medical image analysis: probabilistic and non-probabilistic methods.” Medical Image Analysis (2024): 103223.
July 17	Student Paper Presentations and Course Wrap-Up	Student presentations.

Learning Format

Registration (closed)

Interested in this seminar, please register via email to

References

[1] Litjens, Geert, et al. “A survey on deep learning in medical image analysis.” Medical image analysis 42 (2017): 60-88.
[2] Shen, Dinggang, Guorong Wu, and Heung-Il Suk. “Deep learning in medical image analysis.” Annual review of biomedical engineering 19.1 (2017): 221-248.
[3] Chen, Xuxin, et al. “Recent advances and clinical applications of deep learning in medical image analysis.” Medical image analysis 79 (2022): 102444.
[4] Tajbakhsh, Nima, et al. “Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation.” Medical image analysis 63 (2020):101693.
[5] Wang, Ge, Jong Chul Ye, and Bruno De Man. “Deep learning for tomographic image reconstruction.” Nature machine intelligence 2.12 (2020): 737-748.
[6] Ben Yedder, Hanene, Ben Cardoen, and Ghassan Hamarneh. “Deep learning for biomedical image reconstruction: A survey.” Artificial intelligence review 54.1 (2021): 215-251.
[7] Huang, Ling, et al. “A review of uncertainty quantification in medical image analysis: probabilistic and non-probabilistic methods.” Medical Image Analysis (2024): 103223.
[8] Castro, Daniel C., Ian Walker, and Ben Glocker. “Causality matters in medical imaging.” Nature Communications 11.1 (2020): 3673.
[9] Jones, Charles, et al. “A causal perspective on dataset bias in machine learning for medical imaging.” Nature Machine Intelligence 6.2 (2024): 138-146.
[10] Thirunavukarasu, Arun James, et al. “Large language models in medicine.” Nature medicine 29.8 (2023): 1930-1940.
[11] Zhang, Jingyi, et al. “Vision-language models for vision tasks: A survey.” IEEE Transactions on Pattern Analysis and Machine Intelligence (2024).

CoMI-L — Summer 2026

Mon, 13 Apr 2026 00:00:00 +0000

Lab: Computational Medical Imaging (CoMI)

Lecturers:
Tutors:
Module: MA-INF 2230
Credits: 4 SWS, 9 CP (270h workload)
Programme: M.Sc. Computer Science
Mode: Optional (2. or 3. Semester)
Language: English
Format: Lab (Praktikum), groups of max. 3 students
Frequency: Every year

Description

Possible project topics (chosen or proposed by the instructors) include:

Anatomical Structure Segmentation (e.g., multi-atlas, CNN-based)
Deformable Registration (e.g., classic demons, learning-based registration)
Lesion/Anomaly Detection (e.g., out-of-distribution detection in X-ray)
Radiomics & Quantitative Biomarker Extraction (e.g., texture features, deep radiomics)
Uncertainty Estimation in Deep Learning (e.g., MC-Dropout, Bayesian networks)
Causality & Bias Analysis in Medical Datasets (e.g., domain shift, fairness)

Learning Goals

Technical Skills

Acquire a deep understanding of a selected problem in computational medical imaging: from problem identification to data processing, algorithm design, implementation, and evaluation.
Implement, test, and validate state-of-the-art image-analysis algorithms (e.g., segmentation, registration, reconstruction, classification) on real medical data.
Develop proficiency in using libraries and toolkits commonly used in medical imaging (e.g., ITK, SimpleITK, PyTorch for deep learning, nibabel/dipy).
Learn to document software systematically (API documentation, code comments, unit tests) and to generate reproducible experiments (version control, containerization).

Soft Skills

Collaborate effectively in small groups over the entire semester, dividing tasks (e.g., data curation, algorithmic implementation, evaluation).
Prepare clear, concise technical documentation and design rationales.
Present project progress and final results in oral form (midterm demo, final demo).
Critically assess one’s own results and position them within the state-of-the-art in computational medical imaging.

🗓 Project (Summer 2026)

Segmentation-Aware Benchmarking of Clinical, Radiomic, Deep-Learning and Foundation-Model Biomarkers for NSCLC Survival Prediction on Pretreatment CT

Can AI support reproducible imaging-biomarker discovery for non-small cell lung cancer (NSCLC)? Students benchmark zero-shot and promptable tumour-segmentation models, clinical risk models, handcrafted radiomics and pretrained CT foundation-model representations for prognosis and phenotype discovery. The central question: do prognostic conclusions change when the tumour contour comes from an automated model instead of an expert?

Dataset: — pretreatment thoracic CT of 422 patients with manual 3D gross-tumour-volume contours and survival data ( ), the cohort of the landmark radiomics study by Aerts et al.
Team: 4 students; all share data handling and phenotyping, each leads one experimental work package.
Software: all experiments run on the lab’s open for consistent pre-processing, metrics and figures.

WP	Topic	Lead
1	Data handling, normalisation, harmonisation and quality control	all
2	Zero-shot / promptable segmentation benchmark (MedSAM, SAM-Med3D, FastSAM3D)	S1
3	Clinical survival model (Cox, random survival forest, calibration)	S2
4	Radiomics (PyRadiomics), feature selection and stability under automated contours	S3
5	Foundation-model benchmarking (CT-CLIP, DINOv2, MedicalNet)	S4
6	Correlation analysis, phenotyping and survival curves	all

Learning Format

Students work in groups of up to 3 on a semester-long project. Each group will:

Identify a well-defined problem and collect or be provided a suitable dataset (publicly available or hospital data under ethics approval).
Perform a literature review to select baseline methods.
Implement baseline and at least one novel or improved approach.
Evaluate quantitatively (e.g., Dice, Hausdorff, MSE, classification accuracy) and qualitatively (expert review where possible).
Document the software (README, API docs) and provide a reproducible environment (Docker/Singularity).
Present midterm progress (oral demonstration + short written summary) and final results (oral + poster or full report).

Assessment

Graded Exam:
Oral presentation (final demo + Q&A, 20 minutes per group). Written report (12–16 pages, including background, methods, results, discussion).

Ungraded Coursework (required for admission to the exam):
Project work; attendance in course sessions in accordance with the exam regulations of 2023, § 12(6).

Recommended Prerequisites

At least one of the following:

MA-INF 2222 – Visual Data Analysis
MA-INF 2312 – Image Acquisition and Analysis in Neuroscience
MA-INF 2220 – Lab Visualization and Medical Image Analysis (or equivalent)

References and Resources

Relevant Journals:

Nature Machine Intelligence
Medical Image Analysis (MedIA)
IEEE Transactions on Medical Imaging (TMI)
Radiology: AI, European Radiology, IEEE Journal of Biomedical and Health Informatics

Relevant Conferences:

MICCAI (Medical Image Computing & Computer-Assisted Intervention)
CVPR (IEEE/CVF Computer Vision and Pattern Recognition)
MIDL (Medical Imaging with Deep Learning)
ISBI (International Symposium on Biomedical Imaging)
RSNA (Radiological Society of North America Annual Meeting)

Registration

Registration takes place via eCampus or by contacting .

Federated CT foundation models for multi-center detection of lymph node metastasis in pancreatic cancer

Thu, 09 Apr 2026 00:00:00 +0000

PDAC-CTFM — Scientific Report Paper

Sun, 05 Apr 2026 00:00:00 +0000

Overview

This work introduces a privacy-preserving federated deep learning framework for preoperative lymph node metastasis (LNM) detection in pancreatic ductal adenocarcinoma (PDAC). We integrate large-scale CT foundation model pre-training with a heterogeneity-aware federated aggregation strategy to enable collaborative training across three German institutions without sharing raw patient data.

Key contributions:

Fine-tuning of the on multi-center PDAC CT data for LNM classification.
A heterogeneity-aware federated aggregation strategy (PDAC-CTFMOurs) that jointly accounts for label-distribution imbalance and representation-level divergence across clients.
Multi-center evaluation on 546 patients from three institutions (UKB Bonn, Charité Berlin, UMG Göttingen).

Main results (aggregated test set):

Model	Balanced Acc.	DOR	AUPRC
PDAC-CTFMFedAvg	0.4607	0.2500	0.6968
PDAC-CTFMOurs	0.5866	2.1010	0.6752
PDAC-CTFMCentral (upper bound)	0.6010	3.4541	0.7376

Our federated method outperforms all FL baselines (FedAvg, FedDisco, FedBN, FedALA) by +12.6% balanced accuracy over FedAvg while preserving data privacy.

2. Federated Learning — Proposed Method (Heterogeneity-Aware FedDisco)

Runs PDAC-CTFMOurs with default hyperparameters (a=0.1, b=0.1, γ=10). To sweep hyperparameters, edit the a_values, b_values, gamma_values lists at the bottom of the script:


### Using the code

The repository README covers installation, data, demo data (synthetic), using real data, federated learning setup. See [the repository on GitHub](https://github.com/albarqounilab/PDAC-CTFM#readme) for step-by-step instructions.

AiM — Summer 2026

Mon, 23 Mar 2026 00:00:00 +0000

Block Course: AI in Medicine (AiM)

Lecturers:
Tutors: , , , , ,
Module: CM1659
Credits: 2 SWS
Programme: PhD Candidates at BIGC CPS and NeuroScience
Language: English
Format: Block Course (2 SWS)
Time: Tuesday to Thursday, 14-16 April 2026, 09:00 – 17:00
Location: UKB, Venusberg-Campus 1, building C66, 1st floor, seminar room 101

Description

This block course offers an intensive introduction to artificial intelligence in medicine, with a focus on modern deep learning methods for biomedical and clinical applications. Designed as an interdisciplinary format, the course combines concept-driven lectures with interactive labs, allowing students to engage both with the theoretical foundations of AI and with practical challenges in medical data analysis.

The course is structured around a series of core topics that are central to contemporary AI in medicine. It begins with an introduction to deep learning, providing the conceptual basis for understanding neural networks and their role in medical applications. Building on this foundation, students are introduced to key tasks such as detection and classification, as well as segmentation, which are essential in areas such as diagnostic imaging, image-guided analysis, and automated clinical decision support. The course then expands to quality control in AI systems, with a particular emphasis on uncertainty estimation and interpretability, highlighting how trustworthy and transparent models can be developed and assessed. Further sessions address bias and causality, encouraging students to critically examine fairness, generalizability, and the difference between correlation-driven prediction and clinically meaningful reasoning. Finally, the course introduces vision-language models and their emerging role in connecting medical images, clinical text, and multimodal reasoning.

Each lecture is followed by an interactive lab in which students apply, explore, and critically reflect on the methods discussed. These labs are designed to deepen understanding through hands-on experimentation, collaborative problem-solving, and discussion of real-world medical use cases. In this way, the course not only introduces students to the technical building blocks of AI in medicine but also emphasizes the broader methodological and ethical questions that arise when such systems are developed for healthcare.

By the end of the course, students will have gained a structured overview of important AI methods in medicine, an understanding of their main areas of application, and a critical perspective on their limitations, risks, and clinical potential.

Learning Goals

Technical Skills

Understand and critically appraise current research in AI in Medicine.
Deepen expertise in medical image segmentation, registration, classification, uncertainty quantification, and bias/causality.

Topics and Focus Areas

Seminar topics are drawn from recent publications in top-tier venues and may include:

Deep Learning Architectures (U-Nets, GANs, Vision Transformers)
Segmentation & Classification (e.g., anatomical structure delineation, lesion detection)
Uncertainty Quantification & Bias (Bayesian learning, causal inference, dataset shift)

🗓 Block Course Schedule (14-16 April 2026)

Date	Time	Topic	Description
Tuesday, April 14	09:00 – 09:30	Intro
	09:30 – 10:30	Lecture: Detection and Classification	Introduction to SoTA algorithms for Detection and Classification
	11:00 – 12:30	Lecture: Quality Control: Interpretbility	Introduction to SoTA Interpretability algorithms
	14:00 – 17:00	Lab: Detection and Classification	Introduction to SoTA algorithms for Detection and Classification
Wednesday, April 15	09:00 – 10:30	Lecture: Segmentation	Introduction to deep learning concepts specific to medical imaging.
	11:00 – 12:30	Lecture: Quality Control: Uncertiantity	Introduction to uncertainty and quantification methods
	14:00 – 17:00	Lab: Segmentation and Interpretbility	Introduction to SoTA algorithms for Segmentation and Interpretability
Thursday, April 16	09:00 – 10:30	Lecture: Bias and Causality	Introduction to Bias and Causality in Medical Imaging
	11:00 – 12:30	Lecture: Vision-Language Models	Introduction to Vision-Language Models
	14:00 – 17:00	Lab: Segmentation (Cont.)	Free time to complete the labs.

Recommended Prerequisites

At least one of the following:

Familiarity with fundamental medical imaging techniques and an interest in AI applications.
Understanding of medical diagnostics and clinical data interpretation.
A solid background in Python programming, linear algebra, probability, and numerical algorithms is recommended.

References and Resources

[1] Litjens, Geert, et al. “A survey on deep learning in medical image analysis.” Medical image analysis 42 (2017): 60-88.
[2] Shen, Dinggang, Guorong Wu, and Heung-Il Suk. “Deep learning in medical image analysis.” Annual review of biomedical engineering 19.1 (2017): 221-248.
[3] Chen, Xuxin, et al. “Recent advances and clinical applications of deep learning in medical image analysis.” Medical image analysis 79 (2022): 102444.
[4] Tajbakhsh, Nima, et al. “Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation.” Medical image analysis 63 (2020):101693.
[5] Wang, Ge, Jong Chul Ye, and Bruno De Man. “Deep learning for tomographic image reconstruction.” Nature machine intelligence 2.12 (2020): 737-748.

Registration

Registration takes place via or by contacting .

AiM-BIGS

Mon, 23 Mar 2026 00:00:00 +0000

Using the code

The repository README covers the practical training for aim bigs course module — recycled from the bild summer school ( , datasets. See for step-by-step instructions.

CIR/01/2026: Ph.D. position in AI in Medicine: Foundation Models (m/f/d)

Fri, 13 Mar 2026 00:00:00 +0000

We combine excellence in research, teaching, and patient care. The University Hospital Bonn (UKB) is a maximum care hospital with more than 1,300 beds. With around 38 clinics and 31 institutes as well as more than 8,000 employees (over 5,000 full-time staff), the UKB is one of the largest employers in Bonn. Every year, the UKB treats around 50,000 inpatients and around 35,000 emergencies, as well as provides over 350,000 outpatient treatments.

The following (TVL E13 - 75%) Ph.D. position is available at the Computational Imaging Research (CIR) Lab, headed by Prof. Dr. Shadi Albarqouni, in the Clinic for Diagnostic and Interventional Radiology of the University Hospital Bonn, University of Bonn:

Ph.D. position in AI in Medicine: Foundation Models (m/f/d)

starting April 2026 or as agreed upon. The position is initially limited to three years, with the possibility of extension.

The Ph.D. position will be based in the newly founded research lab for Computational Imaging Research (CIR), which aims to develop i) fully automated, highly accurate innovative computational methods that save expert labor and efforts, and mitigate the challenges in medical imaging; namely the availability of a few annotated data, low inter-/intra-observers agreement, inter-/intra-scanners variability and domain shift, ii) innovative deep Federated Learning algorithms that can fairly distill and share the knowledge among AI agents in a robust and privacy-preserved way, and iii) affordable AI algorithms suitable for low-quality data generated by low-resource settings and point-of-care devices. As AI technology becomes the de facto knowledge discovery approach in many industries, including Healthcare, robustness, and explainability have emerged as key factors to be considered in .

This project focuses on leveraging Foundation Models (FMs) for advanced medical imaging and computational phenotyping. The PhD candidate will develop methods that integrate vision, language, and clinical data to identify clinically meaningful patient phenotypes from heterogeneous and partially annotated datasets.

The research will explore multimodal representation learning, weakly supervised approaches, and the use of FMs for annotation support and knowledge transfer. Emphasis will be placed on building robust, generalizable, and interpretable models for real-world clinical applications, enabling improved patient stratification and precision medicine.

The Ph.D. candidate will have the opportunity to work on cutting-edge research that bridges the gap between advanced natural language processing techniques and medical image analysis. This includes collaborating with clinicians to identify clinical use cases where VLMs can contribute to improved diagnostic workflows, as well as developing methods to ensure the transparency, interpretability, and regulatory compliance of AI models incorporating LLMs in healthcare.

The Ph.D. candidate will be enrolled in the at the University of Bonn. If you have a strong background in deep learning, medical imaging, and an interest in exploring the potential of LLMs within this domain, this position offers a unique opportunity to be at the forefront of AI research in medicine.

Your responsibilities:

Build and create a few clinical use cases for benchmarking existing state-of-the-art (SOTA) algorithms. This includes running baselines and pre-/post-processing pipelines
Develop innovative computational algorithms for Robustness against data and model poisoning attacks, and Explainability of Deep Learning models
Publish and present scientific outcomes at Intl. conferences and high-impact journals
Maintain close collaboration with the team members and clinical partners

Your qualifications:

M.Sc. in Computer Science, Machine Learning, or equivalent with an interest in Medical Imaging
Strong knowledge of Machine/Deep Learning with experience in discriminative models, adversarial attacks, and Bayesian neural networks
Excellent programming skills in Python and including fundamental software engineering principles and machine learning design patterns
Excellent analytical, technical, and problem-solving skills
Be highly motivated and a team player with excellent communication and presentation skills, including experience in communicating across discipline boundaries
Fluent command of the English language

Desirable qualifications:

Track record of publications at top-tier conferences and high-impact journals in the field, e.g., CVPR, MICCAI, MIDL, IEEE TMI and MedIA.
Hands-on experience with the
Working in a Linux environment, with experience in shell and cluster (SLURM) scripting
Fluency in spoken and written German

What we offer you:

A secure future: remuneration according to the German salary scale TV-L (E13 - 75%)
Flexible for families: flexible working time, home office, onsite nursery, and parental care.
Provisions for later: company pension scheme
Discounted public transport ticket: discounted ticket for public transport (VRS) on-site health management service: Numerous health promotion offers
Employer benefits: Discounted offers for employees
Subsidized continuing education and training

The University of Bonn is committed to diversity and equal opportunity and is certified as a family-friendly university. It aims to increase the proportion of women in areas where women are under-represented and to promote their careers in particular. Therefore, we strongly encourage applications from qualified women. Applications will be handled in accordance with the State Equality Act (Landesgleichstellungsgesetz). Applications from individuals with a certified severe disability and from those of equal status are particularly welcome.

Contact:

If you meet the requirements and you are looking for a challenging job? Do not hesitate and send your application including a cover letter (highlighting your qualifications), a detailed CV (with links to previous projects and code), scanned academic degrees, and the contact details of two referees (preferably by e-mail in a single PDF file up to 5 MB in size), quoting the job advertisement no. CIR/01/2026 in your email’s subject to .

Press release on our DAAD SANAD Project -- written by Hanan Nasser from LAU, Lebanon (08.01.2026)

Thu, 08 Jan 2026 00:00:00 +0000

Artificial intelligence for ultrasound-based detection of giant cell arteritis

Thu, 01 Jan 2026 00:00:00 +0000

INTERACT-CMIL: Multi-Task Shared Learning and Inter-Task Consistency for Conjunctival Melanocytic Intraepithelial Lesion Grading

Thu, 01 Jan 2026 00:00:00 +0000

KI-gestützte datenschutzkonforme multizentrische Vorhersage lymphogener Metastasierung beim Pankreaskarzinom mittels eines föderierter CT-Foundation-Modelle

Thu, 01 Jan 2026 00:00:00 +0000

Mapping the integration of artificial intelligence and digital technologies in health technology assessment: a scoping review protocol of global knowledge and practices

Thu, 01 Jan 2026 00:00:00 +0000

ProSona: Prompt-Guided Personalization for Multi-Expert Medical Image Segmentation

Thu, 01 Jan 2026 00:00:00 +0000

Organzie the Symposium on AI in Medicine and Food Sciences

Tue, 09 Dec 2025 00:00:00 +0000

SANAD Symposium on AI in Medicine & Food Sciences

Over the past two days, we had the pleasure of hosting our SANAD symposium on AI in Medicine and Food Sciences within the framework of the DAAD Ta’ziz project, welcoming a delegation from Tunisia, Lebanon, and Iraq.
Project link:

The symposium brought together colleagues from Universitätsklinikum Bonn, The University of Bonn, and our partner universities in the MENA region: Lebanese American University, University of Tunis El Manar, and Duhok Polytechnic University. Together, we focused on strengthening academic cooperation, interdisciplinary exchange, and capacity building under the SANAD vision.

We extend our gratitude to Prof. Björn Krüger for his excellent keynote, to all speakers and participants for their contributions, and to our dedicated team for their support.

Special thanks to Dr. Ismail Acir and his team for co-organizing and hosting the Food Sciences Day with great hospitality.

Beyond the academic program, our delegation enjoyed a rich cultural experience — from the Christmas market in Bonn to the Chocolate Museum and the Cologne Cathedral.

We look forward to building on these connections and continuing our collaboration.

Symposium on AI in Medicine & Food Science

Mon, 01 Dec 2025 00:00:00 +0000

Time: All times correspond to local Bonn time (CET)

Day 1 — Tuesday, 9 December 2025

Arrival & Check-in
📍 Youth Hostel Bonn, Haager Weg 42, 53127 Bonn

Day 2 — Wednesday, 10 December 2025

Symposium on AI in Medicine
📍 Baunscheidtstraße 17, 53113 Bonn

Time	Session	Speaker
09:30–10:00	Registration	—
10:00–10:30	Welcome Speech
10:30–11:45	Keynote Speech: Seeing, Sensing, Securing: AI in Modern Medicine
11:45–12:00	Therapeutic & Surgical Simulation in Medicine
12:00–12:15	AI-Driven Segmentation for Early Eye Disease Detection
12:15–12:30	“Ocean modeling in the Eastern Mediterranean: insights from physical based modelling, data assimilation and machine learning.
12:30–14:00	Lunch Break	—
14:00–14:15	FIDAL: Diversity-Aware Federated Active Learning Under Real-World Distribution Shifts
14:15–14:30	MRIQT: Physics-Aware Diffusion Model for Image Quality Transfer in Neonatal Ultra-Low-Field MRI
14:30–14:45	Rapid and affordable detection of Giant Cell Arteritis using B-mode ultrasound
14:45–15:15	Break	—
15:15–15:30	Teleradiology System Integrating DICOMweb and AI	Molka Yermani
15:30–15:45	Lightweight Bayesian Deep Learning for Point-of-Care Breast Ultrasound	Zeravan Ali
15:45–16:00	Modality‑Aware Deep Learning for Human Bone Fracture Recognition on HBFMID	Omar Al-Zakholi
16:00–16:15	Embedding BI-RADS Foundational Concepts for Cross-Domain and Cross-Modality Generalization in Breast Imaging
16:15–16:30	ML for Histopathological Classification of Conjunctival Melanocytic Intraepithelial Lesion
16:30–16:45	Affordable AI for Retinal Disease Detection
16:45–17:00	Wrap-up Discussion

Day 3 — Thursday, 11 December 2025

Symposium on Food Science
📍 Käthe-Kümmel-Str. 1, 53115 Bonn

Time	Session	Speaker
14:25–14:30	Welcome Speech	&
14:30–15:00	Proteins Under the Microscope: Extraction, Modification and Characterisation
15:00–15:10	Effects of processing on protein-bound amino acids
15:10–15:30	Innovative plant protein sources: Common bean and hempseed press cake as raw material for protein ingredient production	&
15:30–16:00	Break	—
16:00–16:15	From Questions to Discoveries: Highlights from My Food Science Research Career
16:15–16:30	Driving Excellence in Nutrition and Food Science: LAU’s NFS Department’s Mission and Work
16:30–16:40	Micro and Nanoplastics in table salt: occurence and exposure	Hanin El Hajj
16:40–16:50	Unmasking gluten contamination in gluten-free products	Roa Echtay
16:50–17:10	Wrap-up Discussion	Shadi & Ismail
17:10–17:30	Lab Tour

We have the symposium a bit late as one of the key guest’s expertise in Food Science will only arrive in the afternoon

Day 4 — Friday, 12 December 2025

Roundtable discussion in our lab + Excursion Day (Self-Organized)** 📍 Baunscheidtstraße 17, 53113 Bonn

For the Excurssion, here are some suggested visits: Bonn Christmas Market, Cologne Cathedral, Chocolate Museum Köln.
📍 Bonn / Cologne region

Day 5 — Saturday, 13 December 2025

Departure Day
📍 Youth Hostel Bonn

Invited Keynote at the 12th International Conference on Intelligent Computing and Information Systems (ICICIS) in Cairo, Egypt

Tue, 25 Nov 2025 00:00:00 +0000

I had the pleasure to give a keynote on Leveraging Collective Intelligence for Inclusive Global Healthcare at the 12th International Conference on Intelligent Computing and Information Systems (ICICIS) in Cairo, Egypt. I had the chance to meet and discuss further collaboration with the dean at the Ain Shams University Prof. Dr. Rasha Ismail, Prof. Dr. Hassan Abdalla from East London University, and Prof. Robert Sloan, from University of Illinois Chicago, and Prof. Huan Liu from Arizona Satet University.

Contributed to the GLOHRA’s fact-finding mission to Brussels

Mon, 17 Nov 2025 00:00:00 +0000

I recently had the privilege of joining the Steering Committee of the German Alliance for Global Health Research (GLOHRA) in Brussels for a fact-finding mission from 17–20 November 2025. Alongside colleagues from diverse research disciplines, we engaged in constructive dialogue with key European institutions and partners to better understand the EU’s policy and funding landscape for global health research. Over three days of meetings, we gained valuable insights into emerging priorities for research funding and the upcoming EU Framework Programme (FP10), identifying new opportunities for collaboration and coordinated action from the research community. This visit not only strengthened our understanding of how global health research can be supported at the EU level but also reinforced our collective commitment to advancing impactful, interdisciplinary research that addresses global health challenges.

INTERACT-CMIL

Mon, 17 Nov 2025 00:00:00 +0000

About the method

From the abstract of :

Accurate grading of Conjunctival Melanocytic Intraepithelial Lesions (CMIL) is essential for treatment and melanoma prediction but remains difficult due to subtle morphological cues and interrelated diagnostic criteria. We introduce INTERACT-CMIL, a multi-head deep learning framework that jointly predicts five histopathological axes; WHO4, WHO5, horizontal spread, vertical spread, and cytologic atypia, through Shared Feature Learning with Combinatorial Partial Supervision and an Inter-Dependence Loss enforcing crosstask consistency. Trained and evaluated on a newly curated, multicenter dataset of 486 expert-annotated conjunctival biopsy patches from three university hospitals, INTERACT-CMIL achieves consistent improvements over CNN and foundation-model (FM) baselines, with relative macro $\mathrm{F}_{1}$ gains up to $55.1 \%(\text{WHO} 4)$ and 25.0% (vertical spread). The framework provides coherent, interpretable multicriteria predictions aligned with expert grading, offering a reproducible computational benchmark for CMIL diagnosis and a step toward standardized digital ocular pathology.

MRIQT: Physics-Aware Diffusion Model for Image Quality Transfer in Neonatal Ultra-Low-Field MRI

Mon, 17 Nov 2025 00:00:00 +0000

AiR

Mon, 10 Nov 2025 00:00:00 +0000

1. Project Overview

This exercise introduces students to medical image classification using deep learning on Chest X-ray datasets.
Students will explore model performance, generalization, and robustness by training models on a subset of X-rays and evaluating them under controlled challenges such as class imbalance and out-of-distribution (OOD) testing.

Learning goals:

Understand the deep learning workflow for medical image classification.
Use pre-trained foundation models (e.g., ResNet, DenseNet, or ViT from torchvision or timm) for feature extraction.
Handle imbalanced data and evaluate OOD performance.
Develop critical insight into model reliability in medical contexts.

Using the code

The repository README covers 2. dataset. See for step-by-step instructions.

MRIQT

Sun, 02 Nov 2025 00:00:00 +0000

About the method

From the abstract of :

Portable ultra-low-field MRI (uLF-MRI, 0.064 T) offers accessible neuroimaging for neonatal care but suffers from low signal-to-noise ratio and poor diagnostic quality compared to high-field (HF) MRI. We propose MRIQT, a 3D conditional diffusion framework for image quality transfer (IQT) from uLF to HF MRI. MRIQT combines realistic K-space degradation for physics-consistent uLF simulation, v-prediction with classifier-free guidance for stable image-to-image generation, and an SNR-weighted 3D perceptual loss for anatomical fidelity. The model denoises from a noised uLF input conditioned on the same scan, leveraging volumetric attention-UNet architecture for structure-preserving translation. Trained on a neonatal cohort with diverse pathologies, MRIQT surpasses recent GAN and CNN baselines in PSNR 15.3% with 1.78% over the state of the art, while physicians rated 85% of its outputs as good quality with clear pathology present. MRIQT enables high-fidelity, diffusion-based enhancement of portable ultra-low-field (uLF) MRI for deliable neonatal brain assessment.

Using the code

The repository README covers setup, test on a data sample, train on your own dataset, prepare your data in ., generate your own transfer function from your paired data. See for step-by-step instructions.

ProSona

Sun, 02 Nov 2025 00:00:00 +0000

About the method

From the abstract of :

Automated medical image segmentation suffers from high interobserver variability, particularly in tasks such as lung nodule delineation, where experts often disagree. Existing approaches either collapse this variability into a consensus mask or rely on separate model branches for each annotator. We introduce ProSona, a two-stage framework that learns a continuous latent space of annotation styles, enabling controllable personalization via natural language prompts. A probabilistic U-Net backbone captures diverse expert hypotheses, while a prompt-guided projection mechanism navigates this latent space to generate personalized segmentations. A multi-level contrastive objective aligns textual and visual representations, promoting disentangled and interpretable expert styles. Across the LIDC-IDRI lung nodule and multi-institutional prostate MRI datasets, ProSona reduces the Generalized Energy Distance by 17 % and improves mean Dice by more than one point compared with DPersona. These results demonstrate that natural-language prompts can provide flexible, accurate, and interpretable control over personalized medical image segmentation. Our implementation is available online1https://github.com/albarqounilab/ProSona.

Using the code

The repository README covers requirements, quick start (lidc), data, training. See for step-by-step instructions.

David participated with an oral presentation at the 10th BIGS Clinical & Population Science Student Retreat

Mon, 27 Oct 2025 00:00:00 +0000

On 27–28 Oct 2025 in Bonn, I attended the 10th BIGS Clinical & Population Science Student Retreat—a focused two-day program across the life sciences that highlighted post-PhD career pathways and the central role of human-in-the-loop AI in healthcare—featuring inspiring keynotes, the EnaCom “Research-to-Use-Case” workshop, and targeted networking.

In the oral presentation I showcased our work, “Multi-center Collaboration for Improved Lymph Node Metastasis Detection in PDAC.” Building on [1], commenting on the retrospective multicenter cohort of 652 PDAC patients (391 with lymph-node metastases; 59.97%), with the expanded two German sites—University Medicine Berlin (BER) and University Medical Center Göttingen (GÖT)—to include a more rigorously stress-test our two-stage deep-learning pipeline.

[1] David Dueñas Gaviria, Patrick Kupczyk, Belinda Lee, Peter Gibbs, Hyun Ko, Alexander Semaan, and Shadi Albarqouni. Deep learning for lymph node metastasis detection in pancreatic ductal adenocarcinoma. In Ruidan Su, Alejandro F. Frangi, and Yudong Zhang, editors, Proceedings of 2024 International Conference on Medical Imaging and Computer-Aided Diagnosis (MICAD 2024), pages 66–76, Singapore, 2025. Springer Nature Singapore.

AIR — Winter 2025/26

Mon, 13 Oct 2025 00:00:00 +0000

Special Version: AI in Radiology – From Pixels to Decisions (AiR-GUC Edition)

Coordinators: (University of Bonn), & (German University in Cairo)
Tutors:
Duration: October 23, 2025 – February 5, 2026 (every Thursday, 09:00 AM - 11:00 AM)
Format: Blended Seminar + Project-based Collaboration
Credits: 4 CP (Bonn) / Elective (GUC)
Language: English

Course Overview

Artificial Intelligence (AI) is rapidly transforming medical imaging, enhancing clinical workflows and diagnostic accuracy. This special edition of the AiR seminar, titled “AI in Radiology – From Pixels to Decisions,” bridges technical and medical perspectives, offering students a deep dive into the theoretical foundations, applications, and challenges of AI in Radiology.

This joint course brings together:

Medical students from the University of Bonn
Computer Science students from the German University in Cairo (GUC)

It is jointly funded under the and aims to foster international collaboration, interdisciplinary teamwork, and cross-cultural learning through guided lectures, interactive sessions, and project-based research.

Learning Objectives

By the end of this course, students will:

Understand fundamental and advanced AI methods applied to medical imaging.
Analyze academic research papers and translate ideas into practical mini-projects.
Collaborate across disciplines to integrate technical and clinical insights.
Present and communicate interdisciplinary work effectively.
Develop sustainable academic and cultural connections between Germany and Egypt.

Teaching & Learning Format

Hybrid sessions: alternating between live online sessions and optional in-person meetings.
Group projects: each team (2 GUC CS students + 1 Uni Bonn medical student) develops an interdisciplinary mini-project throughout the semester.
Guest lectures: experts from Germany and Egypt provide advanced insights into AI in healthcare.
Final presentations: joint session showcasing team results and reflections.

Seminar Schedule (Winter 2025/26)

Date	Topic	Description	Presenter	Key References / Materials
23 Oct 2025	Kick-off & Orientation	Introduction to the seminar, course goals, and project assignments.	Course Team	—
30 Oct 2025	Foundations of Deep Learning	Overview of deep learning architectures for image analysis.
06 Nov 2025	Detection & Classification in Medical Imaging	Exploring how deep learning identifies and classifies medical abnormalities.
13 Nov 2025	Segmentation in Medical Imaging	Methods for delineating anatomical structures and lesions.
20 Nov 2025	Bias and Causality in AI Models	Understanding dataset bias and causal reasoning in medical AI.		Castro et al., Nat Commun 11 (2020);
27 Nov 2025	Uncertainty Quantification	Quantifying uncertainty and improving trust in AI models.		Huang et al., Med Image Anal (2024).
04 Dec 2025	Foundation Models & Vision-Language Systems	Exploring transformers and multimodal AI in radiology.
11 Dec 2025	Ethics & Regulatory Aspects	Legal, ethical, and social challenges of deploying AI in healthcare.		Topol, E. Nature Medicine 25 (2019).
18 Dec 2025	Midterm Checkpoint & Group Feedback	Group progress presentations and instructor consultation.	All Teams	—
08 Jan 2026	Guest Talks (Germany–Egypt)	Expert talks on AI applications in clinical and technical domains.	Invited Speakers	—
15 Jan 2026	Student Project Presentations I	Group project presentations (part 1).	Teams	—
22 Jan 2026	Student Project Presentations II	Group project presentations (part 2).	Teams	—
05 Feb 2026	Joint Closing & Reflection	Course wrap-up, cross-cultural discussion, and feedback.	Course Team	—

Grading & Evaluation

GUC Students: Elective course – 100% coursework-based (project & presentation).
Uni Bonn Students: Non-graded seminar – certificate of completion based on attendance, participation, and project contribution.

Collaboration Framework

This seminar is part of the GUC–Uni Bonn Joint Teaching Collaboration, coordinated by:

University of Bonn:
German University in Cairo: ,

Supported by the , this course aims to build sustainable academic bridges between Germany and Egypt.

Contact

University of Bonn
– Clinic for Diagnostic and Interventional Radiology, UKB

German University in Cairo (GUC)
,

CoMI-S — Winter 2025/26

Mon, 13 Oct 2025 00:00:00 +0000

Seminar: Computational Medical Imaging (CoMI)

Lecturers:
Module: MA-INF 2318
Credits: 2 SWS, 4 CP (120 h workload)
Programme: M.Sc. Computer Science
Language: English
Format: Seminar (2 SWS)
Time: Thursdays, 14:00 – 16:00
Location: TBD, INF 6+8

Description

The Computational Medical Imaging (CoMI) seminar provides an opportunity for students to explore, understand, and critically evaluate contemporary research in computational medical imaging — from classical algorithms to modern deep learning-based techniques.

Participants analyze recent publications from leading journals and conferences, deepening their understanding of medical image segmentation, registration, classification, uncertainty quantification, and bias/causality in imaging. The seminar emphasizes both technical expertise and scientific communication skills.

Learning Goals

Technical Skills

Understand and critically appraise current research in computational medical imaging.
Deepen expertise in medical image segmentation, registration, classification, uncertainty quantification, and bias/causality.
Extract core contributions from scientific papers and position them in the context of the state-of-the-art.

Soft Skills

Develop competence in independent literature search and paper analysis.
Present complex research effectively (written and oral), with appropriate visual and didactic support.
Engage in critical discussions and defend arguments scientifically.
Practice time management and constructive peer feedback.

Topics and Focus Areas

Seminar topics are drawn from recent publications in top-tier venues and may include:

Segmentation & Classification (e.g., anatomical structure delineation, lesion detection)
Registration & Alignment (e.g., atlas-based or diffeomorphic registration)
Quantitative Imaging & Radiomics (feature extraction, prognostic modeling)
Uncertainty Quantification & Bias (Bayesian learning, causal inference, dataset shift)
Deep Learning Architectures (U-Nets, GANs, Vision Transformers)
Ethical, Regulatory, and Clinical Translation Aspects

🗓 Seminar Schedule (Winter 2025/26)

Date	Theme	Description	Presenter	Key References / Papers
23 Oct 2025	Kick-off & Introduction	Overview of seminar format, expectations, topic assignments.	Course Team	Duncan, J.S. and Ayache, N., 2002. Medical image analysis: Progress over two decades and the challenges ahead. IEEE transactions on pattern analysis and machine intelligence, 22(1), pp.85-106.
30 Oct 2025	Foundations of Computational Medical Imaging	Classical image processing and analysis methods.		Gonzalez & Woods, Digital Image Processing, Ch2-Ch4
06 Nov 2025	Segmentation and Classification	Delineation of anatomical structures and lesion detection.		Gonzalez & Woods, Digital Image Processing, Ch10-Ch12, Litjens et al., Med Image Anal 42 (2017)
13 Nov 2025	Registration and Alignment	Image registration methods: deformable, atlas-based, and deep learning approaches.		Rueckert, D. and Schnabel, J.A., 2010. Medical image registration. In Biomedical image processing (pp. 131-154). Berlin, Heidelberg: Springer Berlin Heidelberg. ; Sotiras et al., IEEE TMI 32(7):1153–1190 (2013) .
20 Nov 2025	Quantitative Imaging & Radiomics	Feature extraction, radiomic biomarkers, and prognostic modeling.		Lambin et al., Eur J Cancer 48 (2012) ; McCague, C., et al. “Introduction to radiomics for a clinical audience.” Clinical Radiology 78.2 (2023): 83-98. .
27 Nov 2025	Deep Learning Architectures I	Convolutional Neural Networks, U-Nets for image segmentation and classification.		Goodfellow, Ian, et al. Deep learning. Vol. 1. No. 2. Cambridge: MIT press, 2016. Ch-9 ; Ronneberger et al., MICCAI 2015 ;
04 Dec 2025	Deep Learning Architectures II	Generative Adversarial Networks, Vision Transformers, and self-supervised learning.		Goodfellow, Ian, et al. Deep learning. Vol. 1. No. 2. Cambridge: MIT press, 2016. Ch-20 ; Goodfellow et al. 2020 , Isola et al., CVPR 2017
11 Dec 2025	Uncertainty Quantification	Probabilistic and non-probabilistic approaches to quantifying model uncertainty.		Huang et al., Med Image Anal (2024) .
18 Dec 2025	Causality and Bias in Medical Imaging	Causal reasoning and dataset bias in deep learning for medical applications.		Castro et al., Nat Commun 11 (2020) ; Mittermaier, Mirja, Marium M. Raza, and Joseph C. Kvedar. “Bias in AI-based models for medical applications: challenges and mitigation strategies.” NPJ Digital Medicine 6.1 (2023): 113. ; Chinta, Sribala Vidyadhari, et al. “AI-Driven Healthcare: A Review on Ensuring Fairness and Mitigating Bias.” arXiv preprint arXiv:2407.19655 (2024).
08 Jan 2026	Ethics & Clinical Translation	Regulatory and ethical challenges of AI in healthcare.		Topol E., Nature Med 25 (2019) ; Lekadir, Karim, et al. “FUTURE-AI: international consensus guideline for trustworthy and deployable artificial intelligence in healthcare.” bmj 388 (2025).
15 Jan 2026	Student Paper Presentations I	Student presentations and peer feedback.	Students	—
22 Jan 2026	Student Paper Presentations II	Continuation of student presentations.	Students	—
29 Jan 2026	Student Paper Presentations III	Final student session and discussion.	Students	—
05 Feb 2026	Course Wrap-Up & Outlook	Summary of insights and future directions in computational medical imaging.	Course Team	—

Learning Format

Students independently select and present recent research papers in computational medical imaging, fostering discussion and critical evaluation within the seminar group.

Group Presentation: A group of 2-3 Students will present a topic and raise some discussions among the students. They will be asked to write a written summary as a lecture note on their topic (15-20 pages).
Individual Presentation: Each individual will present a relevant paper of her/his choice at the end of the semester

Recommended Prerequisites

At least one of the following:

MA-INF 2222 – Visual Data Analysis
MA-INF 2312 – Image Acquisition and Analysis in Neuroscience

A solid background in Python programming, linear algebra, probability, and numerical algorithms is recommended.

References and Resources

Relevant Journals:

Nature Machine Intelligence
Medical Image Analysis (MedIA)
IEEE Transactions on Medical Imaging (TMI)
Radiology: AI, European Radiology, IEEE Journal of Biomedical and Health Informatics

Relevant Conferences:

MICCAI (Medical Image Computing & Computer-Assisted Intervention)
CVPR (IEEE/CVF Computer Vision and Pattern Recognition)
MIDL (Medical Imaging with Deep Learning)
ISBI (International Symposium on Biomedical Imaging)
RSNA (Radiological Society of North America Annual Meeting)

Registration

Registration takes place via eCampus or by contacting .

Organzie the Summer School on Biomedical Imaging with Deep Learning (BiLD)

Mon, 01 Sep 2025 00:00:00 +0000

We have just wrapped up an inspiring week of the BILD Summer School 2025 in Tunis, and I could not be more grateful for the people and partnerships that made this possible.

A heartfelt thank you to the University of Tunis El Manar, represented by the Rector Prof. Moez CHAFRA and his vice rectors, for their warm welcome and strong support of this initiative. Your presence was an honor and an encouragement to all of us.

My sincere gratitude also goes to my co-organizers, Prof. Monia Najar, Bareen Haval, and Prof. Hussein F. HASSAN, PhD, for their trust and collaboration!

I would like to thank all the speakers and lecturers, Prof. Nassir Navab, Prof. Vasileios Belagiannis, Prof. Islem Rekik, Prof. Ezzeddine Zagrouba, Prof. Hanene Boussi, and Dr. Narjes Ben ameur, Dr. Aditya Rastogi, Dr. Lara Marie Reimer, Leon Nissen, and Rached Abdelkhalek for their amazing talks and lectures.

A special thanks to the tutors: David D. Gaviria, Diletta Chiaro, Ph.D., Elyes Farjallah, Hadil AFFES, and Sarra Guirat for preparing and tutoring a couple of exercises and hands-on sessions during the summer school.

Thanks to our participants, who made it from Lebanon, Iraq, Germany (Ulm, Essen, Munich, and Bonn), and multiple cities in Tunisia (Tunis, Bizerte, Sousse, Sfax, and Kebili), for their dedication and curiosity!

Each of you contributed your expertise, energy, and enthusiasm, turning this event into a true success story.

Highlights of the week:

Engaging lectures that bridged theory and real-world applications in biomedical imaging and AI.
Hands-on sessions where students trained, tested, and discussed DL models—exploring both the challenges and opportunities of deep learning in medical imaging.
Special Hands-on Sessions on MONAI Labels from NVIDIA (Rached Abdelkhalek) and Mobile Medical AI App Development (Lara Marie Reimer and Leon Nissen)
Networking events that fostered meaningful connections and future collaborations across countries and disciplines.

This summer school once again underscored the core mission of the knowledge exchange and building sustainable networks.

This won’t happen without the support of our medical faculty at the Universitätsklinikum Bonn, represented by the Dean, Prof. Bernd Weber, and the DAAD German Academic Exchange Service, with funds from the Auswärtiges Amt (Federal Foreign Office) Germany in Germany.

We look forward to seeing the seeds planted in Tunis grow into lasting collaborations and innovations!

Interested to learn more about the experience shared by our participants, feel free to visit the testimonials page here: .

Interested to join us for the upcoming summer/autumn schools, follow our homepage ( ) and LinkedIn

Participated at the AI for Good Global Summit in Geneva, Switzerland

Tue, 08 Jul 2025 00:00:00 +0000

Just returned from my first AI for Good Global Summit in Geneva — an incredible experience!

I was especially inspired by the focus on AI in healthcare:

Reversing diabetes and Parkinson’s with AI-powered stem cell therapies
Real-time, personalized health risk screening using AI
Tackling global health inequities with open, responsible, and equitable AI solutions

This summit reminded me that AI is not just about innovation — it’s about impact. Excited to keep learning and contributing to this global movement

Panelist at Applied Case Studies of AI in Health for WHO EMRO Workshop, Egypt

Wed, 02 Jul 2025 00:00:00 +0000

Honored to join an inspiring panel at the WHO EMRO AI Workshop on Applied Case Studies of AI in Health, alongside brilliant colleagues working at the intersection of AI, public health, and policy. I shared insights from our work at the University of Bonn on building trustworthy and equitable AI models for medical imaging—especially in global health settings where data diversity and clinical applicability are key challenges. Grateful for the engaging discussion on how we can ensure that AI innovations serve not just high-resource settings, but also strengthen health systems worldwide

Invited Talk at the Andalusian Research Institute in Data Science and Computational Intelligence (DASCI), Spain

Wed, 18 Jun 2025 00:00:00 +0000

I had the pleasure to give a talk on Leveraging Collective Intelligence for Inclusive Global Healthcare at the ! Thanks to Prof. for the invitation and great hospitality! The Andalusian Research Institute in Data Science and Computational Intelligence (DaSCI Institute), is a research institute promoted by the universities of Granada (UGR), Jaén (UJA) and Córdoba (UCO).

Contributed to the GLOHRA SC Meeting and GLOHRA Day in Berlin, Germany

Thu, 05 Jun 2025 00:00:00 +0000

At this year’s GLOHRA Steering Committee Meeting and GLOHRA Day, I helped put AI’s real-world value for global health firmly on the agenda. Co-leading the SC round-table with Prof. Charlotte Koehler, I steered debate away from hype toward actionable next steps. Later, at the Bundestag Parliamentary Evening, we participated in an interesting veent on Global Health in LMICs.

The next day, Jana Fehr and I ran an “Ask the Community” session to crowd-source how we should measure impactful AI. Participants mapped indicators for health outcomes, equity, and system sustainability, then agreed to co-develop a mixed-methods framework all projects can plug into—so evidence accumulates across diseases and settings. The experience clarified my role as a bridge between data science, health systems, and policy.

Multi-center Collaboration for Improved Lymph Node Metastasis Detection in Pancreatic Ductal Adenocarcinoma

Tue, 03 Jun 2025 00:00:00 +0000

Poster presentation at Helmholtz AI Conference 2025: AI for Science

Tue, 03 Jun 2025 00:00:00 +0000

From 3–5 June 2025, David attended the Helmholtz AI Conference 2025 (“HAICON25”) at Messe Karlsruhe, Germany. The Helmholtz AI Conference 2025: AI for Science brought together over 600 researchers, industry practitioners, and students from across Europe to explore cutting-edge applications of artificial intelligence in the sciences—featuring inspiring keynotes, focused sessions, poster, and interactive networking events.'

Over three days of full-track sessions, I had the opportunity to present and discuss our work on Multi-center Collaboration for Improved Lymph Node Metastasis Detection in Pancreatic Ductal Adenocarcinoma. Expanding our previous, with the new retrospective multicenter cohort of 652 PDAC patients (391 with lymph node metastases; 59.97 %), we incorporated two additional German data sources; University Medicine Berlin (BER) and University Medical Center Göttingen (GÖT), providing a more rigorous testbed for our two-stage deep learning framework.

While these results underscore the power of combining automated segmentation priors with clinical features for enhanced LNM detection, the center-specific variability also reinforces the necessity of external validation and domain-adaptation strategies to ensure robust, generalizable performance before clinical deployment.

David also engaged in interactive formats such as the AI World Café, and the Pitch & Networking Event, and gain new perspectives from other disciplines which sparked cross-disciplinary interest. Overall, HAICON25 provided an valuable forum for exchanging ideas with peers in both academic and industrial AI, and for exploring new partnerships that will drive our research forward. I am already looking forward to next year’s meeting, and to applying the insights gained toward our ongoing projects in medical imaging and beyond. – David

OP0343 The ClaudicAItio Model: Ultrasound Image Classification in Giant Cell Arteritis by Application of Artificial Intelligence

Sun, 01 Jun 2025 00:00:00 +0000

Summer School on Biomedical Imaging with Deep Learning (BILD)

Sun, 01 Jun 2025 00:00:00 +0000

Contributed to the 23rd BONFOR Symposium in Bonn, Germany

Wed, 07 May 2025 00:00:00 +0000

At the 23rd BONFOR Symposium on Wednesday, 7 May 2025, held in the DZNE Lecture Hall at University Hospital Bonn, Albarqouni Lab actively showcased its current research by delivering concise oral presentations during the afternoon’s Doctoral Promotion Fair. Our team highlighted a diverse portfolio of projects—ranging from advanced image‐analysis pipelines to AI-driven diagnostic tools—and outlined potential MD-thesis topics for interested students, fostering in-depth discussions about methodological approaches and future research directions. By engaging directly with medical students, the lab not only raised awareness of its work but also provided guidance on how to embark on cutting-edge thesis projects within our group, thereby strengthening the pipeline of young clinician-scientists

Invited Lecture at the Charité – Universitätsmedizin Berlin, Germany

Thu, 10 Apr 2025 00:00:00 +0000

As an Invited Lecturer for the Digital Health and Artificial Intelligence module at Charité – Universitätsmedizin Berlin’s Institute of International Health, I delivered a focused lecture on deep learning algorithms and their medical applications—highlighting federated learning—and had the privilege of engaging with a diverse, global cohort of brilliant medical professionals from Germany, other European countries and the U.S., Africa, the Eastern Mediterranean Region, Asia, and South America.

Invited Talk at the 3rd Symposium on International Pathways to Precision Medicine Symposium (P2M Symposium), Rouen, France

Thu, 03 Apr 2025 00:00:00 +0000

I had the honor of delivering an invited talk at the p2m Symposium, where I had the opportunity to share my work and connect with distinguished experts in the field. The event was an enriching platform that allowed me to engage directly with thought leaders, including Ellen GRANT from Harvard, Fouad CHEBIB from Mayo Clinic, and Ivana Dabaj from Rouen. I am particularly grateful to the event chairs, Soumaya and Abdellah from Rouen, whose exceptional leadership and support made this gathering truly memorable.

Shadi is elected as a member at the Global Young Academy (GYA) (31.03.2025)

Mon, 31 Mar 2025 00:00:00 +0000

Participated at the Helmholtz-ELLIS Workshop on Foundation Models in Science

Mon, 17 Mar 2025 00:00:00 +0000

Elodie, David, and Shadi participated at the Helmholtz ELLIS Workshop on Foundation Models which took place in Berlin over two days, bringing together leading researchers and practitioners from academia and industry. The primary focus was to explore the transformative impact of large-scale AI models (\textit{Foundation Models}) across various scientific domains, including medical imaging, astrophysics, tabular data analysis, and beyond.

Invited Talk at the German University in Cairo and the American University in Cairo, Egypt

Tue, 04 Mar 2025 00:00:00 +0000

Just wrapped up two intense and inspiring days in Cairo at the The German University in Cairo and The American University in Cairo (AUC)! 🇪🇬

From invited talks and an engaging workshop to thesis defenses and valuable networking events, it was a fantastic opportunity to exchange ideas and connect with brilliant minds.

A big thank you to the hosts, Dr. Mohammed A.-Megeed Salem and Dr. Seif Eldawlatly and everyone who made this visit so enriching!

Looking forward to future hashtag#collaborations.

Organized the Seminar on Artificial Intelligence in Radiology: Applications and Research

Thu, 06 Feb 2025 00:00:00 +0000

We have just concluded the last session of our seminar course on Artificial Intelligence in Radiology: Applications and Research which was offered for our medical students at the Faculty of Medicine, University Hospital Bonn, who were keen on leveraging AI technologies to enhance their medical practice, specifically within diagnostic imaging"

Fact Finding Mission to Emirates and Qatar

Thu, 23 Jan 2025 00:00:00 +0000

As part of AGYA projects, I have been on a Fact Finding Mission (FFM) to a couple of universities in Emirates and Qatar. I had the pleasure to meet Dr. Henda Mahmoudi from ICBA, Dr. Nisreen Alwan from Abu Dhabi University, and Dr. Mohammed Yaqoubi from MBZUAI in Emirates and Dr. Sami Zaqout, Dr. Susu Zughaier and the dean of the Medical Faculty, Prof. Marwan Abu-Hijleh, at Qatar University in Qatar.

The lab contributes to the Organizing and Program Committees of MICCAI'25

Thu, 16 Jan 2025 00:00:00 +0000

The lab is thrilled to contribute to the organization of in Daejon, South Korea. This conference is particularly relevant to the lab’s research focus, and we are honored to be part of this journey to advance innovation in AI in Healthcare!

Prof. Dr. Shadi Albarqouni is serving on the Outreach Committee, working to foster collaboration and engagement with the community. Dr. Elodie Germani will serve as an Area Chair, participating in managing the scientific review process and ensuring the quality of accepted papers.

An efficient screening of dissolution modulators of Mg alloys: Combining high-throughput multi-well exposure with topographical quantification of volume loss and pitting corrosion

Wed, 01 Jan 2025 00:00:00 +0000

Application of AI and Advanced MRI for a Better Prediction of Disease Progression in Neurological Disorders

Wed, 01 Jan 2025 00:00:00 +0000

TBA

Artery-Fusion Deep Learning for Enhanced Ultrasound Diagnosis of Giant Cell Arteritis

Wed, 01 Jan 2025 00:00:00 +0000

Bias and generalizability of foundation models across datasets in breast mammography

Wed, 01 Jan 2025 00:00:00 +0000

Breast Mammography Dataset (Lebanon & international cohorts)

Wed, 01 Jan 2025 00:00:00 +0000

Partner: Dr. Charbel Mourad. Basis of our MICCAI 2025 study on bias and generalisability of foundation models across mammography datasets.

Bridging Regulatory Science and Medical Imaging Evaluation; and Distributed, Collaborative, and Federated Learning: First International Workshop, BRIDGE 2025, and 6th International Workshop, DeCaF 2025, Held in Conjunction with MICCAI 2025, Daejeon, South Korea, September 23 and September 27, 2025, Proceedings

Wed, 01 Jan 2025 00:00:00 +0000

Conjunctival Melanocytic Intraepithelial Lesion (CMIL) Histopathology Dataset

Wed, 01 Jan 2025 00:00:00 +0000

Curated with Prof. Martina C. Herwig-Carl (Ophthalmic Pathology, UKB) together with University Hospitals Tübingen and Erlangen.

Deep Learning for Lymph Node Metastasis Detection in Pancreatic Ductal Adenocarcinoma

Wed, 01 Jan 2025 00:00:00 +0000

Enhancing patient-specific deep learning based segmentation for abdominal magnetic resonance imaging-guided radiation therapy: a framework conditioned on prior segmentation

Wed, 01 Jan 2025 00:00:00 +0000

Exploring the limit of image resolution for human expert classification of vascular ultrasound images in giant cell arteritis and healthy subjects: the GCA-US-AI project

Wed, 01 Jan 2025 00:00:00 +0000

Fetal Ultrasound Dataset for Congenital Diaphragmatic Hernia

Wed, 01 Jan 2025 00:00:00 +0000

Partner: PD Dr. Florian Recker, Center for Obstetrics and Gynecology, University Hospital Bonn.

FUTURE-AI: International consensus guideline for trustworthy and deployable artificial intelligence in healthcare

Wed, 01 Jan 2025 00:00:00 +0000

Giant Cell Arteritis Ultrasound Dataset

Wed, 01 Jan 2025 00:00:00 +0000

Curated with the Center for Rare Rheumatological Diseases (Claus-Jürgen Bauer, Prof. Valentin Schäfer), University Hospital Bonn. One of the largest clinically validated datasets for ultrasound-based GCA diagnosis; being extended towards a multi-center German cohort.

Our lab led the ML-readiness: quality control, cleaning, standardised preprocessing, annotation workflow and training/validation splits.

Implementation and clinical evaluation of an in-house thoracic auto-segmentation model for 0.35 T magnetic resonance imaging guided radiotherapy

Wed, 01 Jan 2025 00:00:00 +0000

Investigation of the image resolution limit for human interpretation of ultrasound images in giant cell arteritis and healthy subjects-the GCA-US-AI project

Wed, 01 Jan 2025 00:00:00 +0000

Leveraging Collective Intelligence for Inclusive Global Healthcare

Wed, 01 Jan 2025 00:00:00 +0000

LYSTO — Lymphocyte Assessment Hackathon Benchmark

Wed, 01 Jan 2025 00:00:00 +0000

International initiative with Francesco Ciompi (Radboudumc).

MR-Linac Adaptive Radiotherapy MRI Dataset

Wed, 01 Jan 2025 00:00:00 +0000

Partners: Prof. Guillaume Landry, Dr. Christopher Kurz (LMU University Hospital) and Prof. Nassir Navab (TUM), with University Hospital Düsseldorf, Heidelberg, Zurich and Gemelli Rome. Basis of CloverNet and ProSona.

Musculoskeletal Photon-Counting CT Dataset

Wed, 01 Jan 2025 00:00:00 +0000

Acquired at the Department of Radiology, UKB, on photon-counting CT. Used together with labelled CT data from NAIST to study domain shift and federated learning feasibility.

Neonatal Ultra-Low-Field / High-Field MRI Dataset

Wed, 01 Jan 2025 00:00:00 +0000

Curated with the Department of Experimental Neonatology, University Hospital Bonn. Basis of MRIQT, a physics-aware 3D diffusion model for image-quality transfer; expert clinicians rated 85% of generated images as diagnostically useful.

Pancreatic Ductal Adenocarcinoma Multi-Center CT Dataset

Wed, 01 Jan 2025 00:00:00 +0000

Partners: Patrick Kupczyk (Radiology) and Alexander Semaan (Surgery), University Hospital Bonn, with Charité – Universitätsmedizin Berlin and UMG Göttingen. Harmonised across sites with a federated evaluation setup, so that data never leave the hospitals.

Prostate MRI Dataset with MR Fingerprinting

Wed, 01 Jan 2025 00:00:00 +0000

Curated with Dr. Barbara Wichtmann and Prof. Julian Luetkens (Radiology, UKB); extended with an external cohort from the Peter MacCallum Cancer Centre, Melbourne, and public data (PI-CAI).

Retinal Fundus Image Dataset

Wed, 01 Jan 2025 00:00:00 +0000

International collaboration with Dr. Said El Shamieh (Beirut Arab University, Lebanon). Harmonisation, label mapping and bias/generalisation evaluation protocols led by our lab.

Strategic Arab-German Network for Affordable and Democratized AI (SANAD)

Wed, 01 Jan 2025 00:00:00 +0000

In recent years, the transformative potential of Artificial Intelligence (AI) in revolutionizing global healthcare has become evident, offering promising advancements in diagnostic precision, personalized treatment, and healthcare delivery. However, integrating AI into low-resource environments, especially in low- and middle-income countries (LMICs), remains constrained by significant barriers such as limited computational resources, inadequate local data, and financial constraints. The Strategic Arab-German Network for Affordable and Democratized AI in Healthcare (SANAD- سند) seeks to address these challenges by fostering collaboration, empowering local expertise, and advancing Affordable AI and Collaborative Federated Learning solutions tailored to underserved regions. The word “SANAD” means support in Arabic, which is what this project is about.

SANAD aims to bridge the healthcare gap between resource-rich and resource-limited settings, particularly the Global North and Global South disparity. Many existing AI models are developed using centralized datasets from well-funded institutions, which often fail to generalize across diverse populations and contexts. SANAD proposes a strategic, inclusive approach to democratizing access to AI by leveraging interdisciplinary expertise and promoting knowledge-sharing between Arab and German stakeholders

Measures:

, Tunisia
, Germany

Unleashing the power of AI in predicting the technological and phenolic maturity of pomegranates cultivated in Lebanon

Wed, 01 Jan 2025 00:00:00 +0000

Met our partners at Peter MacCallum, Melbourne, Australia

Thu, 19 Dec 2024 00:00:00 +0000

During my recent trip to Melbourne, Australia, supported by the from the University of Bonn, I had the pleasure of meeting with the lab members of and at the . The visit provided fascinating insights into their research themes, ranging from studying zebrafish to Drosophila. I was particularly intrigued by the parallels between these model organisms and the human body, as well as the short life cycle of Drosophila, which facilitates the study of diverse behavioral questions such as jet lag and emotional connections.

We also explored opportunities for collaboration in areas including pancreatic ductal adenocarcinoma (PDAC), breast cancer, and prostate cancer. This visit builds on the strong ties established during the delegation from Bonn to Melbourne in March 2023 ( ), further strengthening the foundation for impactful research collaborations. We have also discussed the closing repoty of our common project on .

Potential Collaboration on organizing a Summer School on AI in Medicine, Tunisia

Tue, 26 Nov 2024 00:00:00 +0000

Last week, I had the privilege of meeting Prof. Dr. Moez CHAFRA, President of the University of Tunis El Manar, and his esteemed colleagues. We had a productive discussion exploring potential collaborations, including the exciting possibility of organizing a Summer School on AI in Medicine in Tunisia, among other initiatives.

Excited to see what the future holds—stay tuned for updates!

Invited Lecture at Health Data Science Principles (GD5301) at St. Andrews University, UK

Thu, 21 Nov 2024 00:00:00 +0000

I had the pleasure to give a lecture at the course offered at St. Andrews University and interact with the briliant students there! Thanks to Prof. from the School of Medicine for the great hospitality!

David participated at MICAD 2024 in Manchester، UK

Tue, 19 Nov 2024 00:00:00 +0000

At MICAD 2024 in Manchester, UK, David presented our research, Deep Learning for Lymph Node Metastasis Detection in Pancreatic Ductal Adenocarcinoma. This study introduces an innovative two-stage deep learning framework to improve the detection of lymph node metastases (LNM) in PDAC. Utilizing 3D CT scans and integrating imaging with clinical features, our model achieves enhanced performance without relying on lymph node segmentation.

Our study underscores the significance of combining imaging data with non-imaging clinical attributes, such as tumor size, anatomical location, and CA-19-9 tumor marker levels. The model demonstrated a 35.32% relative improvement in balanced accuracy, highlighting the impact of holistic data integration in LNM detection. Notably, the inclusion of straightforward automated segmentation priors further optimized performance, marking a pivotal step forward in preoperative diagnostics for PDAC.

Engaging with the MICAD community gave us a valuable platform to showcase these advancements, fostering discussions on multi-institutional collaboration and addressing challenges like domain shifts and data heterogeneity. The feedback and interest from peers emphasized the relevance of our findings, opening avenues for potential collaborations to refine and expand this work.

David participated at the MIB Future Panel in Bonn

Wed, 06 Nov 2024 00:00:00 +0000

It was an exciting to present our research on Lymph Node Metastasis Detection in Pancreatic Ductal Adenocarcinoma at the MIB Future Panel 2024. Hosted by the Medical Imaging Center Bonn in partnership with Universitätsklinikum Bonn and the Medical Faculty of Rheinische Friedrich-Wilhelms-Universität Bonn, this symposium united thought leaders from both academia and industry. The event highlighted the critical importance of multi-center collaborations for the development of robust Computer-Aided Diagnosis (CAD) systems, which are essential for providing radiologists with reliable tools suited for real-world clinical application.

Moderator at the MIB Future Panel 2024

Wed, 06 Nov 2024 00:00:00 +0000

It was a pleasure moderating the Symposium on AI in Medicine and sharing our recent works, which have recently been accepted at hashtag#MICAD and hashtag#MICCAI Workshops! Thanks to the team members Sarah Schaab, David D. Gaviria, and Elyes Farjallah for their great efforts! Thanks to the co-moderator Prof. Valentin S. Schäfer, and the organizing team behind the MIB Future Panel; Lisa Mona Marie Senner, Prof. Frank G. Holz, and the dean, Prof. Bernd Weber, among others! Congrats to the winners!

David participated with a poster presentation at the Student Retreat from BIGS Clinical and Population Science PhD program

Mon, 04 Nov 2024 00:00:00 +0000

David participated in the Student Retreat 2024 organized by the BIGS Clinical and Population Science PhD program in Bonn, Germany. This event was an invaluable opportunity to present my research in a poster format, engage with fellow scholars, and exchange scientific knowledge. The retreat provided a platform to receive constructive feedback on my work, expand my network, and enhance my understanding of diverse research topics within clinical and population science.

Elodie attended and presented at the 2024 International Conference on Image Processing (ICIP) in Abu Dhabi, UAE

Wed, 30 Oct 2024 00:00:00 +0000

Elodie had the chance to attend the 2024 IEEE International Conference on Image Processing in Abu Dhabi, Emirates. She presented her work during an oral session about Medical Image Analysis. This work was part of her Ph.D. projects and was entitled Uncovering communities of pipelines in the task fMRI analytical space. In this work, she used graph representations to explore the relationships between fMRI pipelines results and the stability of their relationships across different study contexts. The results were useful to evaluate the potential generalisability and robustness to unseen context of methods to mitigate analytical variability, for instance, style transfer between statistic maps.

Invited Keynote at the 1st International NEU DESAM Biotechnology Congress in Nicosia

Fri, 18 Oct 2024 00:00:00 +0000

I had the pleasure to give a talk at the DESAM Congress at the Near East University in Nicosia. I had the chance to meet and discuss further collaboration with reprsented by the rector, Prof. Dr. Tamer ŞANLIDAĞ, vice rector; Prof. Dr. Umut AKSOY, and his advisors; Prof. Dr. Murat SAYAN and Assoc. Prof. Dr. Dilber UZUN ÖZŞAHİN at the Near East University.

AIR — Winter 2024/25

Mon, 14 Oct 2024 00:00:00 +0000

Seminar (2 SWS): Artificial Intelligence in Radiology: Applications and Research (Artifizielle Intelligenz in der Radiologie: Anwendungen und Forschung)

Lecturers:

Tutors: ,

Time: November 15 – Jan 31 (every Friday, ~~11:30 AM - 1:00 PM~~ 16:00 - 17:30)

Format: Online (Zoom Meeting) with potential face-to-face meetings, if participants interested! Link was provided via email.

Language: English

eCampus: to eCampus

Description

Course Objectives

Upon completing this seminar, students will be able to:

Grasp deep learning concepts and their specific applications in medical imaging.
Analyze advanced image analysis tasks, including classification, detection, segmentation, and disease progression modeling.
Evaluate ethical and regulatory considerations in the deployment of AI within healthcare.
Discuss the future potential of AI-driven solutions in personalized medicine and patient care.

Target Audience

Prerequisites and Desired Knowledge

Prerequisites: Familiarity with fundamental medical imaging techniques and an interest in AI applications.
Desired: Understanding of medical diagnostics and clinical data interpretation.

Seminar Topics and Schedule (Tentative)

Date	Topic	Description	Materials
Nov 22	Deep Learning Foundations	Introduction to deep learning concepts specific to medical imaging.	Meskó, Bertalan, and Marton Görög. “A short guide for medical professionals in the era of artificial intelligence.” NPJ digital medicine 3.1 (2020): 126.
Nov 29	Detection and Classification in Medical Imaging	Exploring detection and classification techniques in diagnosing diseases from medical images.	Litjens, Geert, et al. “A survey on deep learning in medical image analysis.” Medical image analysis 42 (2017): 60-88. , Chen, Xuxin, et al. “Recent advances and clinical applications of deep learning in medical image analysis.” Medical image analysis 79 (2022): 102444. ,
Dec 6	Segmentation	Deep learning methods for segmenting key structures in medical images.	Tajbakhsh, Nima, et al. “Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation.” Medical image analysis 63 (2020):101693.
Dec 13	Causality and Bias	Causality and Bias in deep learning with medical imaging.	Castro, Daniel C., Ian Walker, and Ben Glocker. “Causality matters in medical imaging.” Nature Communications 11.1 (2020): 3673. , Jones, Charles, et al. “A causal perspective on dataset bias in machine learning for medical imaging.” Nature Machine Intelligence 6.2 (2024): 138-146
Dec 20	~~Uncertainity Quantification~~	Moved to the next year due to holiday season
Jan 10	Q&A Session	Students should prepare questions for their assigned papers and discuss them with the tutors
Jan 17	Uncertainity Quantification	Quanitifying Uncertainty in Deep Learning Models	Huang, Ling, et al. “A review of uncertainty quantification in medical image analysis: probabilistic and non-probabilistic methods.” Medical Image Analysis (2024): 103223.
Jan 24	Student Paper Presentations (1)	Student presentations.
Jan 31	Student Paper Presentations (2) and Course Wrap-Up	Wrap-up, feedback, and discussion on future AI applications in medical practice.

Learning Format

Registration (closed)

Interested in this seminar, please register via email to and

References

[1] Litjens, Geert, et al. “A survey on deep learning in medical image analysis.” Medical image analysis 42 (2017): 60-88.
[2] Shen, Dinggang, Guorong Wu, and Heung-Il Suk. “Deep learning in medical image analysis.” Annual review of biomedical engineering 19.1 (2017): 221-248.
[3] Chen, Xuxin, et al. “Recent advances and clinical applications of deep learning in medical image analysis.” Medical image analysis 79 (2022): 102444.
[4] Tajbakhsh, Nima, et al. “Embracing imperfect datasets: A review of deep learning solutions for medical image segmentation.” Medical image analysis 63 (2020):101693.
[5] Wang, Ge, Jong Chul Ye, and Bruno De Man. “Deep learning for tomographic image reconstruction.” Nature machine intelligence 2.12 (2020): 737-748.
[6] Ben Yedder, Hanene, Ben Cardoen, and Ghassan Hamarneh. “Deep learning for biomedical image reconstruction: A survey.” Artificial intelligence review 54.1 (2021): 215-251.
[7] Huang, Ling, et al. “A review of uncertainty quantification in medical image analysis: probabilistic and non-probabilistic methods.” Medical Image Analysis (2024): 103223.
[8] Castro, Daniel C., Ian Walker, and Ben Glocker. “Causality matters in medical imaging.” Nature Communications 11.1 (2020): 3673.
[9] Jones, Charles, et al. “A causal perspective on dataset bias in machine learning for medical imaging.” Nature Machine Intelligence 6.2 (2024): 138-146.
[10] Thirunavukarasu, Arun James, et al. “Large language models in medicine.” Nature medicine 29.8 (2023): 1930-1940.
[11] Zhang, Jingyi, et al. “Vision-language models for vision tasks: A survey.” IEEE Transactions on Pattern Analysis and Machine Intelligence (2024).

Participate at the World Health Summit in Berlin

Sun, 13 Oct 2024 00:00:00 +0000

At the World Health Summit, I had the opportunity to participate in key sessions moderated by GLOHRA, such as “Stronger Together: Improving Research Partnerships with LMICs,” where we heard from influential figures, including parliament members, and sessions like “Bridging Policy and Research: Translating the EU Global Health Strategy into Action” and “Building Success: On the Road to the ‘Nutrition for Growth’.” A few key takeaways from these discussions included the importance of: i) fostering communication that respects diverse cultures, ii) upholding the 3Es—engagement, equity, and ethics—in our partnerships, and iii) delivering impact and ensuring its sustainability in LMIC research collaborations. Moreover, the focus on nutrition for all, especially through affordable diet programs in LMICs, is critical. Capacity building in nutrition, potentially through AI and technology, also emerged as a vital strategy for driving progress.

Participate at the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI)

Sun, 06 Oct 2024 00:00:00 +0000

Despite the difficult circumstances, we are proud to contribute to MICCAI 2024 with the following papers and events:

Our paper on Affordable DL for common and inherited retinal diseases will be presented at the 11th Ophthalmology Image Analysis (OMIA) Workshop on 10th October, AM Session, Cristal Room - Palmeraie Palace. Further details can be found here: .

with Elyes Farjallah, Razieh Rezaei, Philipp Herrmann, Sandrine Kuenzel, Frank G. Holz from Universitätsklinikum Bonn, and Said El Shamieh from Beirut Arab University.

Our paper on Leveraging Planning annotations in Adaptive Radiation Therapy will be presented at #CLIP Workshop on 6th October, AM Session, Conference Center. Further details can be found here: .

with Francesca De Benetti, Yousef Yaganeh, Claus Belka, Stefanie Corradini, Nassir Navab, Christopher Kurz, Guillaume Landry, and Thomas Wendler

Other exciting events during the conference include:

#RISE Networking Event where Nassir Navab will deliver an inspiring talk on the democratization of healthcare ( .
Our 5th workshop on Distributed, Collaborative, and Federated Learning (#DeCaF) – will be held on 10th October, PM Session ( .

On a personal note, serving as Challenges Co-Chair at #MICCAI2024 has been incredibly difficult, especially with my family being under threat in #Gaza during these troubling times. However, I’ve pushed through, believing that #resilience in the face of adversity will make us stronger.

Looking ahead, I’m optimistic about the future and excited for a productive year filled with new achievements. I’m especially looking forward to meeting and interacting with #young #talents at #MICCAI in #Marrakech next week.

AAI6430 — Winter 2024/25

Tue, 01 Oct 2024 00:00:00 +0000

AI for Medical Diagnosis and Prediction

Lecturers:

Tutors: TBA

Time: March - April 2025 (tentative)

Course Description

This is an 8-week course designed for non-computer science students at the Lebanese American University. The course equips students with practical skills to implement AI in the field of medical diagnosis and prediction. Students will explore techniques like medical image classification, segmentation, object detection, and reconstruction, as well as time-series data analysis for medical forecasting. Additionally, they will learn about weakly-, semi-, and self-supervised learning, and how to tackle fairness and robustness challenges in healthcare AI.

Course Objectives

By the end of this course, students will:

Understand the role of AI in medical diagnosis and prediction.
Apply AI techniques to medical image analysis and time-series forecasting.
Explore weakly-, semi-, and self-supervised learning in healthcare.
Analyze ethical implications, fairness, and robustness in healthcare AI models.
Gain hands-on experience with coding medical AI solutions.

Target Audience

This course is tailored for non-computer science students who have taken introductory courses in Machine Learning and Deep Learning. The students are expected to have an interest in applying AI to medical applications but do not require deep knowledge of computer science concepts.

Course Schedule (12 hours per week)

Week	Lecture (3 hrs)	Description	Practical Sessions (3 hrs)
Week 1	Introduction to AI in Healthcare	Overview of AI in healthcare, challenges, and opportunities.	Image classification using CNNs on medical datasets.
Week 2	Medical Image Segmentation	Explore AI segmentation techniques like U-Net.	Implement U-Net for medical image segmentation.
Week 3	Object Detection in Medical Imaging	Object detection models and challenges in medical images.	Implement YOLO/Faster R-CNN on medical datasets.
Week 4	Medical Image Reconstruction & Time-Series Data	AI techniques for image reconstruction and medical time-series.	Time-series classification using RNN/LSTM models.
Week 5	Time-Series Forecasting & Weakly Supervised Learning	Forecasting techniques and weakly supervised learning in healthcare.	Apply forecasting models and weakly labeled data.
Week 6	Semi-Supervised & Self-Supervised Learning	Explore semi- and self-supervised learning.	Train models with limited labels using contrastive learning.
Week 7	Fairness & Robustness in AI for Healthcare	Ethical AI, fairness, and robustness in healthcare AI models.	Analyze bias and robustness in healthcare datasets.
Week 8	Advanced Applications & Final Project Presentation	Explore advanced AI applications and wrap-up project presentations.	Student presentations and peer review of final projects.

Course Materials

Lecture slides and readings will be provided weekly.
Jupyter notebooks for hands-on labs.

Evaluation and Projects

Students will be evaluated through a combination of:

Predefined Projects: Students will work on predefined medical AI projects, implementing techniques learned throughout the course.
MCQs: Quizzes covering theoretical aspects of AI and its applications in healthcare.
Project Presentations: Final presentations of projects in Week 8.
Peer Reviews: Students will review each other’s projects for additional feedback and learning.

Prerequisites

Students should have a basic understanding of machine learning and deep learning, as well as some experience with Python programming.

Contact Information

For any inquiries, please contact the instructor at .

Invited Lecture for the GLOHRA Academy Series

Tue, 01 Oct 2024 00:00:00 +0000

Join us in our #GLOHRA session on Federated Learning with Medical Imaging! This is quite an important topic nowadays to tackle the data privacy issue while leveraging the collective intelligence we have in multi-national and diverse medical institutes around the globe! It is the way forward to bridge the gap between the global north and global south in medicine and healthcare! This aligns with our recent activities in organizing the which will happen in Morocco in conjunction with MICCAI, and our recent . I am looking forward to meeting you on 1st October!

Invited Keynote at the First Arab Science Symposium in Hannover, Germany

Fri, 20 Sep 2024 00:00:00 +0000

I am deeply grateful to the organizers of this symposium for bringing together Syrian and Arab scholars in such a meaningful way. It is a privilege to have the opportunity to present on the topic of collective intelligence, highlighting the importance of interdisciplinary research in shaping a better future. Events like these foster collaboration and inspire innovative solutions that transcend boundaries, and I am honored to be a part of it.

Participate at the DAAD PRIME Alumni Seminar

Fri, 13 Sep 2024 00:00:00 +0000

I was fortunate enough to attend and participate at the DAAD PRIME Alumni Seminar where I had the chance to meet and re-connect with PRIME fellows and a couple of colleagues! We exchanged career paths and trajectories and heard several success stories in both academia and industries. Although I received the 6 years ago, I felt it was just yesterday! Gratfeul to the DAAD for this fellowship and tremendous efforts to reconnect with our fellows!

Invited Talk at North American Imaging in MS Cooperative, USA

Wed, 11 Sep 2024 00:00:00 +0000

It is a pleasure to accept the invitation to deliver a talk at the The Artificial Intelligence and Machine-Learning (AIMS) working group of the North American Imaging in MS (NAIMS) Cooperative. I’ll present our Nature Machine Intelligence Paper on anomaly detection in MR Brain Imaging which includes MS Lesions.

Organzie the Summer School on Affordbale AI (SAAI -- سعي)

Mon, 22 Jul 2024 00:00:00 +0000

We have just wrapped up an amazing week at the .

As the curtains close on an unforgettable week in Bonn, we reflect on the enriching experiences shared at the AGYA Summer School on Affordable AI. We were thrilled to host 60 guests, including 35 students from nations across the globe; 🇲🇷 Mauritania, 🇲🇦 Morocco, 🇹🇳 Tunisia, 🇩🇿 Algeria, 🇱🇾 Libya, 🇪🇬 Egypt, 🇵🇸 Palestine, 🇱🇧 Lebanon, 🇯🇴 Jordan, 🇸🇾 Syria, 🇮🇶 Iraq, 🇴🇲 Oman, 🇶🇦 Qatar, 🇰🇼 Kuwait, 🇸🇦Saudi, 🇦🇪 UAE, 🇾🇪 Yemen, 🇰🇲 Comoros, 🇩🇪 Germany, 🇷🇺 Russia, 🇮🇹 Italy, 🇺🇦 Ukraine, 🇫🇷 France, and 🇨🇴 Colombia alongside 15 faculty and AGYA members, and 6 dedicated tutors.

🚀 Our diverse participants engaged deeply with innovative AI concepts, learning from esteemed speakers, lecturers, and each other. The collaborative environment fostered not only education but also meaningful connections across cultural lines.

🌐 Social events at iconic locations like the Cologne Cathedral, Drachenburg Castle, and the Haus der Geschichte Museum provided fantastic backdrops for our participants to unwind and network, enhancing their cultural and academic experience.

Thanks to everyone who made this event a huge success—your enthusiasm and insights have truly shaped this summer school into a beacon of learning and international cooperation.

Special Thanks to the Arab-German Young Academy of Sciences and Humanities (AGYA) for the financial support. I would like to thank The University of Bonn, Universitätsklinikum Bonn, Helmholtz AI for the host and financial support as well.

Invited Talk at the Summer School for Surgical Data Science

Wed, 17 Jul 2024 00:00:00 +0000

Very much enjoyed the time in Strasbourg!

Great keynotes and lectures at the . I was fortunate enough to meet a couple of brilliant minds at the school! It was also a pleasure meeting my mentor Prof. Nassir Navab!

Thanks for the invitation, organziation and the warm hospitality, Prof. Nicolas Padoy 🙏🙏

Participate at the International Conference on Medical Imaging with Deep Learning (MIDL)

Wed, 03 Jul 2024 00:00:00 +0000

Happy to see so many brilliant works at the hashtag#MIDL conference in Paris! It’s a fantastic opportunity to connect with colleagues and graduates passionate about medical imaging and deep learning.

Invited Keynote Lecture on AI in Ophthalmology and Pathology

Sat, 15 Jun 2024 00:00:00 +0000

I was invited to give a Keynote Lecture on AI in Ophthalmology and Pathology to present a few works on Histopathology; Mitotoc figure detection in Breast Cancer, Domain Adaptation from H&E Imaging to IHC imaging among others. Thanks Prof. Martina C. Herwig-Carl for the invitation and warm hospitality!

Participate at the Helmholtz AI Conference, Germany

Wed, 12 Jun 2024 00:00:00 +0000

We have just participated at the Helmholtz AI conference and presented one of our Helmholtz AI projects!

MR Fingerprinting and Apparent Diffusion Coefficient Mapping for Quantitative Characterization of Clinically Significant Prostate Cancer

Sat, 04 May 2024 00:00:00 +0000

Summer School on Affordable AI (SAAI)

Wed, 01 May 2024 00:00:00 +0000

Research Visit at the University of Sharjah, UAE

Sun, 14 Apr 2024 00:00:00 +0000

I have just completed the research visit at the University of Sharjah, NYU Abu Dhabi, and other esteemed universities. This initiative aims to strengthen connections and foster collaboration between German and Emirati academic institutes in the field of health and medical sciences. I will soon return to the Emirates to organize a one-day workshop on AI and healthcare at a prestigious academic institute. Stay tuned for more details!

Invited to a Panel Discussion on AI in clinical routine at LMU, Munich, Germany

Fri, 19 Jan 2024 00:00:00 +0000

I was excited to join a distinguished panel, including Prof. Dr. Wolfgang Böcker, Prof. Dr. Clemens Cyran, and Katharina Danhauser, moderated by Claus Belka, to explore AI’s readiness for clinical routine. Our discussion spotlighted crucial aspects: the current state of AI in clinics and strategic placement within workflows. Emphasis was placed on robust data governance and addressing development challenges. The dialogue highlighted the importance of collaboration with stakeholders, involving regulatory bodies, and prioritizing education and training. Grateful for the opportunity to contribute to this pivotal conversation shaping the future of AI in healthcare. Thanks for the invitation, Guillaume Landry!

Affordable Deep Learning for Diagnosing Inherited and Common Retinal Diseases via Color Fundus Photography

Mon, 01 Jan 2024 00:00:00 +0000

Bias-aware foundation models for equitable breast cancer screening

Mon, 01 Jan 2024 00:00:00 +0000

This project addresses fairness and generalization challenges of foundation models for breast mammography, where AI systems trained predominantly on data from high-income countries often underperform in underrepresented populations. Together with international clinical partners, we assembled and analyzed a heterogeneous collection of mammography datasets from Europe, North America, the Middle East, and Asia, including a biopsy-confirmed in-house cohort of approximately 1,500 patients from Lebanon. Methodologically, we conducted a systematic evaluation of foundation models under cross-dataset shift and demonstrated that naïve dataset aggregation improves average performance but does not eliminate subgroup-specific biases. Fairness-aware optimization strategies substantially reduced performance disparities across age groups, breast density categories, and geographic regions while maintaining competitive diagnostic accuracy.

Perspective

The collaboration constitutes an ongoing international research effort at the intersection of medical imaging, fairness-aware AI, and global health. Building on current results, the project is being expanded to additional clinical partners in Palestine and Egypt, supported by the Global One Health Research Academy (GLOHRA), and is planned to scale further through the submitted EU Horizon Europe project SHE-Scan, contingent on funding.

CloverNet--Leveraging Planning Annotations for Enhanced Procedural MR Segmentation: An Application to Adaptive Radiation Therapy

Mon, 01 Jan 2024 00:00:00 +0000

Editorial for the Special Issue on the 2022 Medical Imaging with Deep Learning Conference

Mon, 01 Jan 2024 00:00:00 +0000

INTERACT-CMIL: Interpretable multi-task grading of conjunctival melanocytic lesions

Mon, 01 Jan 2024 00:00:00 +0000

Collaborators: University Hospital Tübingen, University Hospital Erlangen

This project addresses the grading of conjunctival melanocytic intraepithelial lesions (CMIL), rare precursors of conjunctival melanoma characterized by complex, interdependent histopathological criteria and substantial inter-observer variability. Together with clinical partners, we curated a multi-center, expert-annotated histopathology dataset comprising 486 conjunctival biopsy image patches labeled along five clinically defined grading axes. Methodologically, we developed INTERACT-CMIL, a multi-task deep learning framework modeling inter-criterion dependencies via selective multi-head supervision and an inter-task consistency loss.

The proposed approach outperformed CNN and pathology foundation-model baselines (macro-F1 up to 85%), producing coherent and interpretable predictions aligned with expert grading standards. Methodological development and evaluation were coordinated within my group in close collaboration with the clinical partners.

Perspective

The project established a reproducible computational benchmark for CMIL grading and contributed to the standardization of digital ocular pathology. A related follow-up project on AI-assisted eyelid tumor analysis received a University of Bonn Prototype Grant, supporting translational development and clinical feasibility studies.

MRIQT: Image-quality transfer from ultra-low-field to high-field neonatal MRI

Mon, 01 Jan 2024 00:00:00 +0000

This project addresses the limited diagnostic usability of portable ultra-low-field MRI (0.064 T) in neonatal neuroimaging, where reduced signal-to-noise ratio constrains clinical interpretation despite bedside accessibility. Together with the clinical partners, we curated a unique in-house cohort comprising 50 paired ultra-low-field and high-field 3D neonatal brain MRI scans, complemented by 230 additional unpaired scans across nine pathologies. Building on this paired–unpaired data structure, we developed MRIQT, a 3D conditional diffusion framework for structure-preserving image quality transfer from ultra-low-field to high-field MRI. The method integrates physics-consistent k-space degradation, classifier-free guided diffusion, and an SNR-weighted volumetric perceptual loss.

MRIQT outperformed CNN- and GAN-based baselines, achieving up to 15.3% improvement in PSNR, while expert clinicians rated 85% of generated images as diagnostically useful with preserved pathology.

Perspective

The project provides a methodological foundation for automated MRI quality enhancement and modality-agnostic quality control. Building on the current results, the work is prepared for translational follow-up and prospective validation, with a planned submission to the Go-Bio Initial funding program.

Robust Federated Learning Methodology for Automatic Chest X-Ray Diagnosis

Mon, 01 Jan 2024 00:00:00 +0000

Robust surgical tool segmentation with state-space vision models

Mon, 01 Jan 2024 00:00:00 +0000

This exploratory collaboration investigated state-space vision models for robust surgical tool segmentation under realistic visual corruptions. In collaboration with Kyung Hee University, we participated in the MICCAI EndoVis 2024 SegSTRONG-C challenge, benchmarking robustness to non-adversarial corruptions such as smoke, bleeding, and illumination changes. The proposed Vision Mamba–UNet approach achieved competitive performance across corruption types and ranked 7th overall among international submissions.

Perspective

The project supported early exploration of state-space models in medical image analysis and international doctoral mentorship. A jointly submitted DAAD exchange proposal, although not funded, laid the groundwork for continued collaboration and future joint research activities.

Two-Steps Approach for Breast Cancer Detection and Classification Using Convolutional Neural Networks.

Mon, 01 Jan 2024 00:00:00 +0000

Organize the Autumn School on AI (EEDA - ايدا)

Thu, 14 Dec 2023 00:00:00 +0000

Thrilled to share the success of our EEDA Autumn School on AI – a global collaboration between Albarqouni Lab (Germany), Universitätsklinikum Bonn, The University of Bonn, and the Faculty of Health Sciences at Beirut Arab University (Lebanon). Over the weekend, a distinguished Lebanese delegation from the Beirut Arab University ( BAU), American University of Beirut ( AUB), and Lebanese American University ( LAU) joined us for an immersive program blending cutting-edge AI with diverse disciplines.

Bridging Gaps, Fostering Excellence: This initiative, supported by the DAAD German Academic Exchange Service Ta’ziz Short Measures Grant, goes beyond an event – a testament to global collaboration. Empowering minds and breaking barriers, we achieved an 80% female participation rate and a competitive 26% acceptance rate, highlighting our commitment to academic excellence.

International Collaboration in Action: Thanks to DAAD, the funding from the Auswärtiges Amt (Federal Foreign Office) Germany, and the endorsement of Arab-German Young Academy of Sciences and Humanities (AGYA), the EEDA Autumn School on AI has sparked meaningful collaborations, from curriculum integration to potential joint Master’s Programs. We are creating a ripple effect of international cooperation and academic brilliance.

Beyond Academia: Our journey surpassed the academic realm, fostering a profound cultural exchange. It’s not just about AI; it’s about building a global community that values diversity, mutual respect, and inclusive collaboration. The delegation enjoyed the social events from visiting the Christmas Market, Cathedral Kolner Döm, to the Haus der Geschichte.

Despite the challenging times, let’s continue breaking barriers, fostering innovation, and creating a better, more connected world. The success of this weekend is just the beginning – together, we can rise!

Photo credits to Paul Nuhn, Universitätsklinikum Bonn

Invited Talk at the Sino-German Symposium on AI in Medicine

Sat, 11 Nov 2023 00:00:00 +0000

Very much enjoyed the time in Shanghai!

Great keynotes and lectures at the . I was fortunate enough to meet a couple of brilliant minds at the school! It was also a pleasure meeting my mentor Prof. Nassir Navab!

Thanks for the invitation, organziation and the warm hospitality, Prof. Kuangyu Shi 🙏🙏

Semi-Supervised Learning for Prostate Lesion Detection and Segmentation

Wed, 01 Nov 2023 00:00:00 +0000

LYSTO: The lymphocyte assessment hackathon and benchmark dataset

Wed, 25 Oct 2023 00:00:00 +0000

David attended ICCV 2023 in Paris

Mon, 02 Oct 2023 00:00:00 +0000

David attended the International Conference on Computer Vision (ICCV) 2023 in Paris from October 2–6, 2023. This event brought together leading minds in computer vision and AI from all over the world. It was an incredible experience to immerse myself in the latest developments in the field and gain insight into cutting-edge research.

The conference covered a wide range of topics, from foundational deep learning techniques to applications in healthcare, autonomous systems, and beyond. The keynote talks were inspiring, and the technical sessions and poster presentations provided valuable ideas and perspectives. It was a fantastic opportunity to connect with the research community and stay up-to-date with the newest trends shaping computer vision.

Affordable AI and Collaborative Federated Learning for Global Healthcare (EEDA)

Fri, 01 Sep 2023 00:00:00 +0000

Invited Talk at German Jordanian University (GIU), Amman, Jordan

Thu, 27 Jul 2023 00:00:00 +0000

IEEE EMBS Jordan Chapter is pleased to invite faculty, students, fresh graduates, and professionals to attend the (virtual) EMBS Jordan Chapter Distinguished Lecture on Wednesday, July 26th, 2023, 6:00-7:30 PM (Jordan Local Time). This time we have a special guest PROF. Dr. Shadi Albarqouni, Professor of Computational Medical Imaging at the University of Bonn in Bonn, Germany. For registration, use the form below or the QR code. The seminar will take place on the ZOOM platform. Details for how to join the virtual meeting will be sent later for those who confirm their registration.

David participates at the AI4Health Summer School

Sat, 03 Jun 2023 00:00:00 +0000

David attended the AI4Health Summer School 2023, organized collaboratively by the French Health Data Hub and French AI research institutions at PariSanté Campus, was an incredible experience. The program focused on advanced methods in medical AI, covering critical aspects of deep learning for medical imaging. I actively participated in two workshops:

Deep Learning for Medical Image Processing: Exploring innovative techniques for tumor segmentation and detection in whole-body 3D MRIs, focusing on patch-based approaches, denoising, contrast correction, and augmentations.
Fed-BioMed Workshop: Examining federated learning approaches tailored for biomedical applications, addressing challenges in privacy-preserving data sharing across institutions.

The event was a truly remarkable opportunity to delve deeper into these cutting-edge topics and engage with peers and leaders in medical AI.

Ascertaining the pose of an x-ray unit relative to an object on the basis of a digital model of the object

Mon, 01 May 2023 00:00:00 +0000

Federated foundation models for multi-center CT staging of pancreatic cancer

Mon, 01 May 2023 00:00:00 +0000

Collaborators: University Medical Center Göttingen, Charité – Universitätsmedizin Berlin, Peter MacCallum Cancer Centre

This project addresses preoperative staging of pancreatic ductal adenocarcinoma (PDAC), where lymph node metastasis is a major prognostic factor but CT-based assessment suffers from low sensitivity and high inter-observer variability. Together with partners across three German centers, we established a retrospective multi-center cohort of 546 patients to enable evaluation under real-world heterogeneity while preserving data privacy. Methodologically, we combined CT foundation model transfer learning (pre-training on 148,000 volumetric CT scans) with heterogeneity-aware federated optimization. The resulting system improved federated performance by +12.6% balanced accuracy over FedAvg and recovered most of the centralized reference performance (centralized BA 0.601; DOR 3.45).

Perspective

The project establishes a sustainable multi-center collaboration for privacy-preserving oncologic imaging in Germany. Ongoing work extends the framework toward multimodal integration and more robust harmonization strategies, and a joint DFG proposal with the clinical partners has been submitted to further advance federated foundation models for PDAC staging and translation.

Deep Learning-based Quantification of Aortic Macroangiopathy Improves Prediction of Cardiovascular Mortality in the National Lung Screening Trial (NLST)

Wed, 01 Mar 2023 00:00:00 +0000

Fact Finding Mission to the United Arab Emirates

Sun, 19 Feb 2023 00:00:00 +0000

Just wrapped up an incredible 72-hour agya fact-finding mission with the brilliant Vice-Presidents and Provosts of top universities in Dubai and Abu Dhabi, UAE. The main goal of our fact-finding mission is to promote the call for membership of the Arab-German Young Academy of Sciences and Humanities (AGYA)!

Dr. Nada El Darra and I had the pleasure to meet Vice-President Sultan Al Haji, the Acting Provost Prof. Tim Baldwin at MBZUAI (Mohamed bin Zayed University of Artificial Intelligence), and the Acting Provost and Chief Academic Officer Prof. Michael Allen at Zayed University besides our brilliant colleagues Dr. Rania Dghaim, Dr. Fatme Al Anouti, and Dr. Omar Alfandi.

Thanks to our alumna, Dr. Henda Mahmoudi, and her colleagues Nour El Jundi and Evans Kioko, we very much enjoyed the wonderful field trip at International Center for Biosaline Agriculture (ICBA) where we discussed potential interdisciplinary research collaboration on AI and Food Safety!

Invited Talk at 3MT Competition at Beirut Arab University

Fri, 10 Feb 2023 00:00:00 +0000

Today I was privileged to give a keynote speech at the Beirut Arab University in Lebanon 🇱🇧. My talk was about the impact of AI in the field of Medicine and how AI is changing the way healthcare professionals diagnose, treat, and care for hashtag#patients. My presentation included a couple of examples from our hashtag#research on microscopic imaging, but I also looked at a couple of hashtag#startups that are taking this to the next level by transforming healthcare through their hashtag#innovative products. Having a fruitful discussion with the audience was a real pleasure for me! Thanks so much, Dr. Nada El Darra and Dr. Said El Shamieh, for the warm welcome and the invitation 🙏

Deep learning for clinically significant prostate cancer in 3D mpMRI and MR Fingerprinting

Sun, 01 Jan 2023 00:00:00 +0000

Collaborators: Peter MacCallum Cancer Centre, Melbourne

This project addresses the detection and characterization of clinically significant prostate cancer (csPCa) in MRI, where diagnostic performance is limited by inter-reader variability and the complexity of multi-parametric, volumetric imaging. Together with the clinical partners, we curated an initial in-house 3D MRI cohort of 106 patients with histopathological reference, including bi-parametric MRI and quantitative MR Fingerprinting (MRF), which was subsequently extended with external and public datasets to form a multi-center cohort of approximately 500 cases. Methodologically, we developed fully 3D deep learning models that explicitly incorporate anatomical prostate zone information, enabling interpretable and biomarker-aware csPCa detection and segmentation beyond slice-wise analysis.

A dedicated sub-project evaluated transformer-based 3D U-Net architectures for lesion segmentation. On public (PI-CAI) and in-house datasets, transformer-based models achieved Dice scores above 0.60 on detected lesions and demonstrated improved robustness under reduced training data compared to CNN-based baselines, while highlighting persistent challenges in false-positive reduction.

Perspective

The project established a scalable research line on interpretable, anatomy-aware AI for prostate cancer imaging and contributed to early-career training through a completed Bachelor’s thesis and an ISMRM abstract. Ongoing work extends this line toward vision–language foundation models integrating 3D imaging, quantitative MRI biomarkers, and anatomical priors, forming a basis for continued multi-center collaboration and future third-party funding.

Deep learning to estimate aging from chest imaging

Sun, 01 Jan 2023 00:00:00 +0000

Chronologic age is an important risk factor for many chronic diseases. However, as we all age at different rates chronological age is an imperfect measure of aging and more accurate estimates of one’s true biological age are desirable. The proposed study will develop and test artificial intelligence-based measures of aging from chest imaging data and will explore whether these measures can improve chronologic age-based clinical guidelines for prevention of cardiovascular disease and cancer. Over recent decades life expectancy has continuously increased, however, the years of life without significant morbidity have remained stable over the last 5 decades resulting in significant socioeconomic burden to society and the healthcare system. Therefore, delaying the onset of morbidity is a promising way to improve healthy aging. In particular, prevention and early detection of diseases of high socioeconomic importance such as cancer and cardiovascular disease are a priority to increase health and longevity. Biological age is a concept to estimate the differences in rates of aging not captured by chronologic age. Several measures of biological age have been developed in recent years including blood, functional and physiological approaches. However, there relevance in clinical routine is limited. Medical imaging may be an effective way to estimate aging by measuring age-related changes based on anatomical alterations visible in the image such as degenerative changes of the spine, dilation of the heart and vasculature and changes in lung parenchyma. Unlike molecular measures, image-based reflections of aging can be calculated opportunistically using data acquired during routine clinical care. Furthermore, these measures can estimate aging of specific organ systems, which may improve disease-specific preventive measures over general estimates of biological age. The image-based estimates of aging using deep learning proposed in this study will address this unmet need and may help to personalize clinical decision-making for chronic disease screening and prevention. My preliminary data show that deep learning can estimate a chest x-ray age from a chest radiograph image and that this chest x-ray age predicts longevity better than chronologic age in large multicenter clinical trial data. The proposed project will build on these promising initial findings with the potential to broaden the application to other imaging modalities and to leverage multi-omics data to infer causes of aging.

Collaborators:

PD. Dr. , UniKlinikum Freiburg

Outcomes

This exploratory project examined the feasibility of imaging-based phenotyping and biological aging estimation from chest CT and MRI. The work focused on conceptual development and clinical hypothesis generation rather than large-scale validation and resulted in a clinical abstract presented at the European Congress of Radiology.

Perspective

The project provided methodological insights into opportunistic imaging-based risk modeling and informed the design of subsequent projects with clearer data availability, validation strategies, and funding perspectives.

Digital staining of white blood cells with confidence estimation

Sun, 01 Jan 2023 00:00:00 +0000

Enhancing oncological care: A guide to setting up a new multidisciplinary cancer cachexia clinic within a tertiary centre

Sun, 01 Jan 2023 00:00:00 +0000

Evolving the era of 5D ultrasound? A systematic literature review on the applications for artificial intelligence ultrasound imaging in obstetrics and gynecology

Sun, 01 Jan 2023 00:00:00 +0000

Guest Editorial Special Issue on Federated Learning for Medical Imaging: Enabling Collaborative Development of Robust AI Models

Sun, 01 Jan 2023 00:00:00 +0000

Joint self-supervised image-volume representation learning with intra-inter contrastive clustering

Sun, 01 Jan 2023 00:00:00 +0000

Lvm-med: Learning large-scale self-supervised vision models for medical imaging via second-order graph matching

Sun, 01 Jan 2023 00:00:00 +0000

Medical Image Computing and Computer Assisted Intervention--MICCAI 2023 Workshops: ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Vancouver, BC, Canada, October 8--12, 2023, Proceedings

Sun, 01 Jan 2023 00:00:00 +0000

Open Medical Inference (OMI)

Sun, 01 Jan 2023 00:00:00 +0000

Outcomes

The Open Medical Inference (OMI) initiative is a BMFTR-funded national research infrastructure aimed at improving reproducibility, transparency, and comparability of machine learning methods in medical imaging and clinical AI. Within OMI, I contributed to the development and evaluation of standardized benchmarking protocols, open evaluation pipelines, and reference datasets for medical AI, with a particular focus on robustness, generalization, and fair comparison across institutions and data distributions. These activities complement my research on federated learning and trustworthy AI by addressing methodological gaps in evaluation practices and supporting community-wide standards for medical machine learning research.

Our role

Contributing researcher (Scientific Lead) with responsibility for methodological input on benchmarking design, evaluation protocols, and reproducibility aspects, in close collaboration with national partners.

Perspective

OMI provides a sustained national framework for open and reproducible medical AI research. The experience and infrastructure developed within OMI inform ongoing and future projects on federated learning, multi-center evaluation, and trustworthy AI, and strengthen national coordination across academic and clinical institutions. These activities align with and complement my work on federated learning benchmarks and multi-center evaluation (see also Sections and ).

Transformer based models for unsupervised anomaly segmentation in brain MR images

Sun, 01 Jan 2023 00:00:00 +0000

Utilising Artificial Intelligence in Cancer Imaging To Improve Patient Outcomes

Sun, 01 Jan 2023 00:00:00 +0000

Cachexia is a devastating tumour-induced wasting disease affecting 30% of all cancerpatients. This initiative brings together clinical, computational (artifi cial intelligence),and laboratory research partners to work on the delivery of improved outcomes forpatients with cancer cachexia. Aims of the collaboration: Funding will provideopportunities to broaden digital research links and submit joint grant and conjoint PhDproposals.

Long-term objectives:

1. Developing non-invasive screening tools for cancercachexia, using AI-assisted imaging analysis approaches.
1. Developing a greaterunderstanding of molecular processes of cancer cachexia using animal models andpatient samples.
1. Increase knowledge and awareness about cachexia througheducational events for clinicians, patients and families on cachexia diagnosis andtreatment.

This multidisciplinary and diverse initiative builds on the expertise andexisting collaborations between researchers/clinicians across both universities (Bonn/Melbourne). The group consists of experts in radiology/ image analysis (Ko,Attenberger), computational science /AI (Albarqouni) and cancer biology/ metabolism(Cheng). This team of clinicians and researchers with deep knowledge of the fi eld ofcancer cachexia are ideally placed to educate clinicians, patients and families oncachexia and increase its awareness in the general community.

Partners:

, Australia
, Germany

Outcomes

This collaboration was supported by institutional funding to foster international exchange at the intersection of artificial intelligence, cancer imaging, and translational oncology. Core outcomes included two jointly organized international workshops held in Bonn and Melbourne, reciprocal research visits, and the identification of shared research directions. The collaboration further supported concrete synergies in ongoing projects, including pancreatic cancer staging and multiparametric prostate MRI.

Perspective

The project established a durable international partnership between the University of Bonn and the University of Melbourne, providing a platform for continued joint research, doctoral training, and future collaborative funding applications.

Fact Finding Mission to Amman, Jordan

Wed, 21 Dec 2022 00:00:00 +0000

Just arrived from Amman, Jordan! It was just an amazing trip to Amman where I met our clinical partner King Hussein Cancer Foundation and Center for one of our projects supported by the Arab-German Young Academy of Sciences and Humanities (AGYA)! I was also happy to accept the invitation of my former PhD student, Tariq Bdair, and the head of his department, Dr. Serin Atiani, to deliver a talk at Princess Sumaya University for Technology! I was impressed by the level of their students, indeed! I would like also to thank Dr. Omar Al-Kadi from The University of Jordan for the invited talk and the great hospitality!

Co-Organizer of the Introduction of Bioinformatics to Arab Education System

Mon, 21 Nov 2022 00:00:00 +0000

I had the pleasure last week of co-organizing a workshop on Bioinformatics in Tunisia. The workshop was part of the working group activities supported by the Arab-German Young Academy of Sciences and Humanities! We had a great lineup of speakers from Germany/Switzerland and Arab countries! I had the pleasure to know a couple of Arab scholars from Egypt, Lebanon, Algeria, Tunisia, and Sudan.

Federated learning for musculoskeletal segmentation in CT

Tue, 01 Nov 2022 00:00:00 +0000

This collaboration explored federated learning for musculoskeletal segmentation in CT imaging across institutions in Germany and Japan. Using labeled CT data from NAIST and an unlabeled whole-body CT cohort from University Hospital Bonn, we evaluated cross-site generalization and domain transfer under realistic scanner and population shifts. In parallel, a secure federated learning infrastructure was deployed and ethical approval obtained at University Hospital Bonn. While the project did not progress to publication-ready federated training due to annotation constraints and deployment complexity, it provided valuable methodological and organizational insights.

Perspective

The collaboration strengthened institutional expertise in federated learning under real-world clinical constraints and informed subsequent national and international projects with more harmonized data availability.

Invited Talk at the Workshop on Collaborative Learning: From Theory to Practice

Sat, 08 Oct 2022 00:00:00 +0000

I had the pleasure to give an invited talk at the #Collaborative Learning workshop at MBZUAI (Mohamed bin Zayed University of Artificial Intelligence)! It was a wonderful weekend full of amazing talks and fruitful discussions! I had the pleasure to meet a few familiar faces in our community along with other great speakers from UC Berkeley, Harvard, MIT, KAUST, ETH Zurich, Nvidia, and EPFL, among others. I would like to thank Michael I. Jordan and the organizing team behind the workshop for the invitation and the excellent hospitality! For those who are interested in the talks, they will be made publicly available soon!

Federated disentangled representation learning for unsupervised brain anomaly detection

Sat, 01 Oct 2022 00:00:00 +0000

Organizing the 2nd version of the MICCAI Workshop on aFfordable AI and healthcare

Thu, 23 Jun 2022 00:00:00 +0000

Affordable deep learning for inherited and common retinal diseases

Wed, 01 Jun 2022 00:00:00 +0000

This international collaboration addresses the diagnosis of inherited retinal diseases (IRDs) and common retinal disorders using affordable color fundus photography, motivated by limited access to genetic testing and advanced imaging in low-resource settings. Together with partners in Germany and Lebanon, we curated and analyzed heterogeneous multi-center retinal imaging datasets combining in-house cohorts and public data, spanning diverse demographics and disease spectra. Methodologically, we developed lightweight deep learning models for multi-class retinal disease classification covering 21 retinal conditions, including 13 IRDs. The approach achieved strong diagnostic performance for prevalent IRDs (e.g., F1 score of 0.86 for retinitis pigmentosa) while remaining computationally efficient and interpretable.

Perspective

The collaboration established a sustainable international research link between ophthalmology groups in Germany and Lebanon and contributed to interdisciplinary student training. Ongoing work explores privacy-preserving and federated learning strategies to improve robustness and equitable deployment across populations further.

Deep learning for fetal diaphragmatic hernia detection in ultrasound

Wed, 01 Jun 2022 00:00:00 +0000

Congenital diaphragmatic hernia (CDH) is a rare but severe fetal malformation associated with high neonatal morbidity and mortality, where early detection is critical for risk stratification, parental counseling, and perinatal management. Although ultrasound is the primary imaging modality in prenatal screening, CDH detection remains challenging due to limited acoustic windows, overlapping anatomical structures, and strong operator dependence.

The initial objective of this project was to develop deep learning–based methods for automated CDH detection in fetal ultrasound. Due to changes in the clinical collaboration context, the project could not be continued in its originally planned data-driven form. As a result, the focus shifted toward consolidating the existing evidence base and systematically analyzing methodological and clinical barriers for AI-assisted fetal ultrasound diagnostics.

As a central outcome, we co-authored a comprehensive systematic review on artificial intelligence in obstetric and gynecologic ultrasound, critically assessing methodological rigor, data availability, validation practices, and clinical readiness across 189 published studies. The review identified fetal malformation detection, including CDH, as a clinically high-impact yet underexplored application area, highlighting the need for standardized datasets, multi-center validation, and interpretable AI approaches.

Perspective

The insights gained from this project provided a methodological foundation for subsequent ultrasound- and MRI-based research with sustained clinical partnerships and informed the design of later projects with improved data availability and validation strategies. Building on this groundwork, renewed efforts toward CDH detection are envisaged within a national multi-center collaboration, leveraging harmonized data collection and shared clinical expertise.

FedDis-NMI — The source code of FedDis

Wed, 01 Jun 2022 00:00:00 +0000

About the method

From the abstract of :

Recent advances in Deep Learning (DL) and the increased use of brain MRI have provided a great opportunity and interest in automated anomaly segmentation to support human interpretation and improve clinical workflow. However, medical imaging must be curated by trained clinicians, which is time-consuming and expensive. Further, data is often scattered across multiple institutions, with privacy regulations limiting its access. Here, we present FedDis (Federated Disentangled representation learning for unsupervised brain pathology segmentation) to collaboratively train an unsupervised deep convolutional neural network on 1532 healthy MR scans from four different institutions, and evaluate its performance in identifying abnormal brain MRIs including multiple sclerosis (MS), vascular lesions, low-grade tumors (LGG), and high-grade tumors/glioblastoma (HGG/GB) on a total of ~538 scans from 6 different institutions and datasets. To mitigate the statistical heterogeneity between the different institutes, we disentangle the parameter space into global, i.e., shape and local, i.e., appearance. We train the shape parameters jointly from four institutes to learn a global model of the healthy anatomical brain structure. The appearance parameters are trained locally on every institute and allow for personalization of the global domain-invariant features with client-specific information, such as scanner or acquisition parameter. We have shown that our collaborative approach, FedDis, improves anomaly segmentation results by 99.74% for MS, 83.33% for vascular lesions,and 40.45% for tumors over locally trained models without the need for annotations or sharing private local data. We found out that FedDis is especially beneficial for clients that share both healthy and anomaly data coming from the same institute, improving their local anomaly detection performance by up to 227% for MS lesions and 77% for brain tumors.

Using the code

The repository README covers requirements, data, usage. See for step-by-step instructions.

Determining a pose of an object in the surroundings of the object by means of multi-task learning

Tue, 01 Feb 2022 00:00:00 +0000

Co-Organizing the 10. DFG-#Nachwuchsakademie

Sun, 23 Jan 2022 00:00:00 +0000

Anomaly-aware multiple instance learning for rare anemia disorder classification

Sat, 01 Jan 2022 00:00:00 +0000

Development of a Deep Learning Toolkit for MRI-Guided Online Adaptive Radiotherapy

Sat, 01 Jan 2022 00:00:00 +0000

In Germany close to 500,000 new patients are diagnosed with cancer every year. More than half of the patients undergo radiotherapy as part of their treatment. In-room imaging, in combination with modern external beam radiotherapy techniques, enables tight adjustment of the delivered dose to the target. However, the full potential of these techniques is currently not exploited. One of the reasons for this is the presence of anatomical changes. For the vast majority of patients in today’s radiotherapy, anatomical alterations are considered by introducing safety margins around the tumor volume during treatment planning, which increases the irradiated volume, the dose burden to organs at risk (OARs), and eventually limits the applicable dose.

An improved treatment can be realized by online adaptive radiation therapy (ART): instead of applying the same irradiation plan throughout the entire course of treatment, the treatment is optimized at each irradiation session on basis of the daily anatomy in treatment position, as inferred from in-room imaging. Great efforts have been made to realize the integration of magnetic resonance imaging (MRI) as in-room imaging modality, but only during the last few years integrated MR-linacs became clinically available at few academic institutions. The superior soft-tissue contrast allows for accurate visualization of targets and OARs, which has allowed basic online ART workflows to be implemented. The time overhead associated with the generation of a synthetic-CT from MRI and a re-delineation of the patient for treatment plan optimization greatly limits applicability to all patients.

The goal of this project is to tackle the main bottlenecks of online plan adaptation: synthetic CT generation and organ delineation. By combining our respective expertise in radiotherapy and artificial intelligence, we aim at proposing deep-learning-based solutions to build a toolkit to tackle both of these challenges. In the case of synthetic CT generation, we will use generative adversarial networks with hybrid paired-unpaired training paradigms, additionally exploiting novel embedding loss functions to ensure geometric fidelity. Alternatively, disentangled representations will be investigated. For segmentation we aim at exploiting the information from previous fractions by performing spatio-temporal segmentation using longitudinal spatial memory networks. Finally, we aim at evaluating the benefits from these approaches by performing an evaluation of their clinical impact in terms of plan optimization, including physician grading of delineations and robustness assessment. The time benefits, including potential corrections to segmentations and synthetic CT, will also be quantified. Further details can be found .

Collaborators:

Prof. , Department of Radiation Oncology, LMU University Hospital
Dr. , Department of Radiation Oncology, LMU University Hospital
Prof. , CAMP, CIT School, TU Munich

Outcomes

MRI-guided online adaptive radiotherapy requires fast and reliable image segmentation and synthesis to support treatment adaptation under strict clinical time constraints. In this DFG-funded project, we addressed key methodological challenges related to patient-specific segmentation and uncertainty-aware modeling under inter-observer and inter-institutional variability. Together with the partners, we curated a longitudinal multi-center MRI cohort comprising planning and fraction scans across multiple institutions, enabling systematic evaluation under realistic clinical heterogeneity. Methodologically, the project resulted in two complementary contributions: CloverNet, a patient-specific 3D segmentation architecture leveraging prior planning annotations, and ProSona, a prompt-guided personalization framework modeling inter-expert variability via a continuous latent space. The proposed methods achieved a significant improvement over standard baselines and received a Best Paper Award at a MICCAI workshop.

Perspective

The project established a strong national research network at the interface of medical imaging, radiation oncology, and AI, and contributed to doctoral and master’s training. Ongoing work extends these approaches toward scalable foundation models for personalized and uncertainty-aware radiotherapy workflows, with a focus on clinical integration and multi-center validation.

FedNorm: Modality-Based Normalization in Federated Learning for Multi-Modal Liver Segmentation

Sat, 01 Jan 2022 00:00:00 +0000

Flamby: Datasets and benchmarks for cross-silo federated learning in realistic healthcare settings

Sat, 01 Jan 2022 00:00:00 +0000

GCA-US-AI: Deep learning–based ultrasound diagnosis of Giant Cell Arteritis

Sat, 01 Jan 2022 00:00:00 +0000

In close interdisciplinary cooperation with clinical rheumatology, this project addresses the highly operator-dependent and time-critical ultrasound diagnosis of Giant Cell Arteritis (GCA), a vasculitis associated with irreversible vision loss. Together with our partners, we curated and annotated an in-house dataset of 3,800 ultrasound examinations from 344 patients, constituting one of the largest clinically validated datasets for ultrasound-based GCA diagnosis. Methodologically, we developed artery-aware deep learning models that integrate anatomical context into ultrasound image analysis, achieving robust performance improvements over artery-agnostic baselines (AUROC up to ~0.9 for larger arteries). Uncertainty estimation and visual explainability were incorporated to support reliable clinical interpretation.

Perspective

The collaboration is ongoing and continues to serve as a basis for joint publications, supervised theses, and continued dataset extension (in-house already ~15,000 images) toward a multi-center cohort. Current work explores vision–language models integrating imaging and clinical data and forms the basis of a planned joint DFG proposal coordinated together with the clinical partners.

Large vision models for robust and trustworthy medical image understanding

Sat, 01 Jan 2022 00:00:00 +0000

This collaboration investigates the applicability of large vision models to medical imaging, addressing robustness, generalization, and trustworthiness under domain shift and limited supervision. The work examined the behavior of foundation-scale models transferred from natural-image pretraining to medical imaging tasks. Methodological contributions include empirical analyses of cross-domain generalization, sensitivity to spurious correlations, and limitations of naïve scaling, alongside proposed evaluation protocols and architectural adaptations to improve robustness and reliability in medical contexts.

Perspective

The collaboration supported doctoral mentorship and strengthened national expertise at the intersection of foundation models and medical imaging. The insights gained inform ongoing work on principled evaluation and responsible deployment of large vision models in medical applications.

ROAM: Random layer mixup for semi-supervised learning in medical images

Sat, 01 Jan 2022 00:00:00 +0000

Semi-Supervised Federated Peer Learning for Skin Lesion Classification

Sat, 01 Jan 2022 00:00:00 +0000

SPA: Shape-Prior Variational Autoencoders for Unsupervised Brain Pathology Segmentation

Sat, 01 Jan 2022 00:00:00 +0000

TriMix: Virtual embeddings and self-consistency for self-supervised learning

Sat, 01 Jan 2022 00:00:00 +0000

Unsupervised Cross-Domain Feature Extraction for Single Blood Cell Image Classification

Sat, 01 Jan 2022 00:00:00 +0000

What can we learn about a generated image corrupting its latent representation?

Sat, 01 Jan 2022 00:00:00 +0000

Organizing a workshop on the Next Generation of AI in Medicine

Tue, 30 Nov 2021 00:00:00 +0000

Co-Organizing the workshop and training on Improving Job Market Skills of Graduates

Sun, 19 Sep 2021 00:00:00 +0000

Organizing the 1st MICCAI Workshop on aFfordable AI and healthcare

Wed, 23 Jun 2021 00:00:00 +0000

BigPicture Project

Mon, 01 Feb 2021 00:00:00 +0000

A new consortium of the EU Innovative Medicines Initiative (IMI) will establish the biggest database of pathology images to accelerate the development of artificial intelligence in medicine.

To take AI development in pathology to the next level, a European consortium combining leading European research centres, hospitals as well as major pharmaceutical industries, is going to develop a repository for the sharing of pathology data. The 6-year project called , will herald a new era in pathology.

Background

Pathology is the cornerstone of the workup of many diseases such as cancer, autoimmune diseases, of the follow up after transplantation and is also critical for the evaluation of the safety of drugs. It’s based on the examination of tissue samples (slides) under the microscope. However, despite its pivotal role, it still relies heavily on the qualitative interpretation by a qualified pathologist.

While the microscope symbolizes the profession, the digitalisation of slides in recent years ignited a revolution: not only images can now be shared and accessed from distant locations, they can also be processed by computers. This opens the door for artificial intelligence (AI) applications to assist the pathologist and help study diseases, find better treatments and contribute to the 3Rs (replace, reduce, and refine animal use in research). However, the development of robust AI applications requires large amounts of data, which in the case of pathology means a huge collection of digital slides and the medical data necessary for their interpretation. Sharing these has so far remained challenging due to the data storage capacity required to host a sufficiently large collection and to concerns regarding the confidential character of the medical information.

To allow the fast development of AI in pathology, the project aims to create the first European, ethical and GDPR-compliant (General Data Protection Regulation), quality-controlled platform, in which both large-scale data and AI algorithms will coexist. The platform will be developed in a sustainable and inclusive way by connecting communities of pathologists, researchers, AI developers, patients, and industry parties.

TU Munich

In this project, , and from , together with will be leading and contributing to the development of Federated Deep Learning algorithms leveraging massive amounts of data, distributed in multiple sources, in a privacy-preserved fashion. This will enable deep learning models to be trained using sensitive data that cannot be made publicly available due to GPDR or sensitivity, e.g. rare diseases. Please visit the website of for further details.

Intended results

The project is divided into four main aspects that concern the large-scale collection of data. First, an infrastructure (hardware and software) must be created to store, share and process millions of images that can be gigabytes each. Second, legal and ethical constraints must be put in place to ensure adequate usage of data while fully respecting patient’s privacy and data confidentiality. Then, an initial set of 3 million digital slides from humans and laboratory animals will be collected and stored into the repository to provide data for the development of pathology AI tools. Finally, functionalities that aid the use of the database as well as the processing of images for diagnostic and research purposes will be developed.

Autoencoders for Unsupervised Anomaly Segmentation in Brain MR Images: A Comparative Study

Fri, 01 Jan 2021 00:00:00 +0000

Butterfly-Net: Spatial-Temporal Architecture For Medical Image Segmentation

Fri, 01 Jan 2021 00:00:00 +0000

Domain Adaptation and Representation Transfer and Distributed and Collaborative Learning: Third MICCAI Workshop, DART 2021, and First MICCAI Workshop, FAIR 2021, Held in Conjunction with MICCAI 2021, Strasbourg, France, September 27 and October 1, 2021: Proceedings

Fri, 01 Jan 2021 00:00:00 +0000

Domain adaptation and representation transfer, and affordable healthcare and AI for resource diverse global health

Fri, 01 Jan 2021 00:00:00 +0000

Eine computergestützte automatische Polypencharakterisierung von Hyperplasten, Adenomen und Serratierten Adenomen im Kolorektum-Ergebnisse der CASSANDRA Studie

Fri, 01 Jan 2021 00:00:00 +0000

Eyesight to AI: Discovery of efficient corrosion modulators via predictive machine learning

Fri, 01 Jan 2021 00:00:00 +0000

Outcome publication (peer-reviewed):

FedDis: Disentangled Federated Learning for Unsupervised Brain Pathology Segmentation

Fri, 01 Jan 2021 00:00:00 +0000

FedPerl: Semi-Supervised Peer Learning for Skin Lesion Classification

Fri, 01 Jan 2021 00:00:00 +0000

Fourier Transform of Percoll Gradients Boosts CNN Classification of Hereditary Hemolytic Anemias

Fri, 01 Jan 2021 00:00:00 +0000

Interpretable learning for rare disease classification in computational hematology

Fri, 01 Jan 2021 00:00:00 +0000

This research line addresses automated analysis of hematological disorders from microscopic blood images, where rare diseases, limited annotations, and cross-site variability challenge conventional deep learning approaches. Using multi-center microscopy datasets covering several rare anemia subtypes, we developed interpretable and data-efficient learning methods, including anomaly-aware multiple instance learning and unsupervised cross-domain feature extraction. The proposed approaches achieved improved generalization and instance-level interpretability compared to supervised and attention-based baselines.

Perspective

Initiated during my appointment at Helmholtz Munich and continued after transitioning to the University of Bonn, this collaboration established a durable national research partnership and contributed to doctoral training. The methodological insights continue to inform ongoing work on robust and weakly supervised learning for rare diseases.

Modeling Healthy Anatomy with Artificial Intelligence for Unsupervised Anomaly Detection in Brain MRI

Fri, 01 Jan 2021 00:00:00 +0000

Semi-Supervised Few-Shot Learning with Prototypical Random Walks

Fri, 01 Jan 2021 00:00:00 +0000

Sickle Cell Disease Severity Prediction from Percoll Gradient Images using Graph Convolutional Networks

Fri, 01 Jan 2021 00:00:00 +0000

The Federated Tumor Segmentation (FeTS) Challenge

Fri, 01 Jan 2021 00:00:00 +0000

The OOD Blind Spot of Unsupervised Anomaly Detection

Fri, 01 Jan 2021 00:00:00 +0000

Multi-task multi-domain learning for digital staining and classification of leukocytes

Mon, 21 Dec 2020 00:00:00 +0000

An Uncertainty-Driven GCN Refinement Strategy for Organ Segmentation

Tue, 01 Dec 2020 00:00:00 +0000

The Future of Digital Health with Federated Learning

Mon, 14 Sep 2020 00:00:00 +0000

6D Camera Relocalization in Ambiguous Scenes via Continuous Multimodal Inference

Sat, 01 Aug 2020 00:00:00 +0000

Fairness by Learning Orthogonal Disentangled Representations

Sat, 01 Aug 2020 00:00:00 +0000

FedDis: Federated disentangled representation learning for brain anomaly detection

Mon, 01 Jun 2020 00:00:00 +0000

This research line addressed one of the central methodological challenges in collaborative medical AI: how to learn clinically meaningful representations across institutions without sharing patient data and without relying on pixel-level pathology annotations. In the context of brain MRI, unsupervised anomaly detection is particularly attractive because lesion annotation is expensive, heterogeneous, and often unavailable, while privacy and governance constraints limit centralized multi-site learning.

Together with the collaborators, we developed FedDis, a federated disentangled representation-learning framework for unsupervised brain anomaly detection. The key methodological idea was to separate the representation space into global shape components, learned collaboratively across institutions, and local appearance components, trained locally to capture client-specific intensity and scanner characteristics. This disentanglement strategy was designed to address statistical heterogeneity across sites while preserving the privacy-preserving benefits of federated learning.

The resulting system was trained collaboratively on 1,532 healthy MRI scans from four institutions and evaluated on 538 pathological volumes from six institutions, covering multiple sclerosis, vascular lesions, and brain tumors. Compared with locally trained unsupervised models, the approach achieved substantial improvements in anomaly segmentation performance, including marked gains for multiple sclerosis, vascular lesions, and tumors, while requiring neither pathology annotations nor raw data sharing. The work was published in Nature Machine Intelligence and became one of the most visible methodological outputs of my collaborative research in federated medical imaging.

Perspective

This project established an important methodological foundation for my later work on federated and collaborative learning, benchmark development, and multi-institutional validation in medical imaging. Conceptually, it shaped my subsequent research agenda by demonstrating that privacy-preserving learning across hospital networks must address not only optimization and communication constraints, but also representation-level heterogeneity across institutions. The insights from this project continue to inform my current and planned work on collaborative validation, foundation models under distribution shift, and trustworthy medical AI infrastructures such as OMI.

Deep Federated Learning in Healthcare

Wed, 01 Apr 2020 00:00:00 +0000

Deep Learning (DL) has emerged as a leading technology for accomplishing many challenging tasks showing outstanding performance in a broad range of computer vision and medical applications. However, this success comes at the cost of collecting and processing a massive amount of data, which often are not accessible, in Healthcare, due to privacy issues. Federated Learning (FL) has been recently introduced to allow training DL models without sharing the data. Instead, DL models at local hubs, i.e. hospitals, share only the trained parameters with a centralized DL model, which is, in return, responsible for updating the local DL models as well.

Our golas in this project is to develop novel models and algorithms for a ground-breaking new generation of deep FL, which can distill the knowledge from local hubs, i.e. hospitals, and edges, i.e. wearable devices, to provide personalized healthcare services.

The principal challenges, to overcome, concern the nature of medical data, namely data heterogeneity; severe class-imbalance, few amounts of annotated data, inter-/intra-scanners variability (domain shift), inter-/intra-observer variability (noisy annotations), system heterogeneity, and privacy issues (see the example below).

Collaboration:

Organizing Committee Member at MICCAI DART 2020

Wed, 01 Apr 2020 00:00:00 +0000

Organizing Committee Member at MICCAI DCL 2020

Wed, 01 Apr 2020 00:00:00 +0000

Uncertainty-based graph convolutional networks for organ segmentation refinement

Wed, 01 Apr 2020 00:00:00 +0000

Seamless Virtual Whole Slide Image Synthesis and Validation Using Perceptual Embedding Consistency

Wed, 19 Feb 2020 00:00:00 +0000

Modelling Labels Uncertainty in Medical Imaging

Wed, 15 Jan 2020 00:00:00 +0000

& Cardoso, MJ (2020)

Wed, 01 Jan 2020 00:00:00 +0000

A learning without forgetting approach to incorporate artifact knowledge in polyp localization tasks

Wed, 01 Jan 2020 00:00:00 +0000

An objective comparison of detection and segmentation algorithms for artefacts in clinical endoscopy

Wed, 01 Jan 2020 00:00:00 +0000

Attention Based Multiple Instance Learning for Classification of Blood Cell Disorders

Wed, 01 Jan 2020 00:00:00 +0000

Bayesian Skip-Autoencoders for Unsupervised Hyperintense Anomaly Detection in High Resolution Brain Mri

Wed, 01 Jan 2020 00:00:00 +0000

Benefit of dual energy CT for lesion localization and classification with convolutional neural networks

Wed, 01 Jan 2020 00:00:00 +0000

Digital staining with confidence estimation for hematological imaging

Wed, 01 Jan 2020 00:00:00 +0000

Collaborators: PD Dr.\ Slobodan Ilic (Siemens AG)

This project investigated the reliability of generative image translation for digital staining of hematological microscopy images, aiming to reduce variability and processing time in laboratory workflows while preserving diagnostically relevant structures. Using paired unstained and chemically stained microscopy images from 24 patients, we developed a style–structure disentangled generative framework with an associated confidence estimation that quantifies output reliability via latent perturbations. The confidence score correlated strongly with perceptual image quality and enabled the identification of unreliable synthetic samples.

Our role

Contributing collaborator with responsibility for methodological development and uncertainty estimation; doctoral supervision and mentorship conducted in close collaboration with academic and industrial partners.

Perspective

The project contributed to the successful doctoral training of an external PhD candidate and established methodological foundations for later work on uncertainty estimation and trust in medical image generation. It exemplifies sustained national collaboration between academia and industry.

Domain Adaptation and Representation Transfer, and Distributed and Collaborative Learning

Wed, 01 Jan 2020 00:00:00 +0000

GANs for medical image analysis

Wed, 01 Jan 2020 00:00:00 +0000

Image-to-Images Translation for Multi-Task Organ Segmentation and Bone Suppression in Chest X-Ray Radiography

Wed, 01 Jan 2020 00:00:00 +0000

Liver lesion localisation and classification with convolutional neural networks: a comparison between conventional and spectral computed tomography

Wed, 01 Jan 2020 00:00:00 +0000

Microaneurysms segmentation and diabetic retinopathy detection by learning discriminative representations

Wed, 01 Jan 2020 00:00:00 +0000

Modelling Uncertainty in Deep Learning for Medical Applications

Wed, 01 Jan 2020 00:00:00 +0000

Deep Learning has emerged as a leading technology for accomplishing many challenging tasks showing outstanding performance in a broad range of applications in computer vision and medical applications. Despite its success and merit in recent state-of-the-art methods, DL tools still lack in robustness hindering its adoption in medical applications. Modeling uncertainty, through Bayesian Inference and Monte-Carlo dropout, has been successfully introduced to computer vision for better understanding the underlying deep learning models. In this proposal, we investigate modeling the uncertainty for medical applications given the well-known challenges in medical image analysis, namely severe class-imbalance, few amounts of labeled data, domain shift, and noisy annotations.

Collaboration:

Prof. , , .

Prof. , ,

Polyp-artifact relationship analysis using graph inductive learned representations

Wed, 01 Jan 2020 00:00:00 +0000

Precise proximal femur fracture classification for interactive training and surgical planning.

Wed, 01 Jan 2020 00:00:00 +0000

Retinal Layer Segmentation Reformulated as OCT Language Processing

Wed, 01 Jan 2020 00:00:00 +0000

Scale-Space Autoencoders for Unsupervised Anomaly Segmentation in Brain MRI

Wed, 01 Jan 2020 00:00:00 +0000

SteGANomaly: Inhibiting CycleGAN Steganography for Unsupervised Anomaly Detection in Brain MRI

Wed, 01 Jan 2020 00:00:00 +0000

Uncertainty Aware Methods for Camera Pose Estimation and Relocalization

Wed, 01 Jan 2020 00:00:00 +0000

Camera pose estimation is the term for determining the 6-DoF rotation and translation parameters of a camera. It is now a key technology in enabling multitudes of applications such as augmented reality, autonomous driving, human computer interaction and robot guidance. For decades, vision scholars have worked on finding the unique solution of this problem. Yet, this trend is witnessing a fundamental change. The recent school of thought has begun to admit that for our highly complex and ambiguous real environments, obtaining a single solution is not sufficient. This has led to a paradigm shift towards estimating rather a range of solutions in the form of full probability or at least explaining the uncertainty of camera pose estimates. Thanks to the advances in Artificial Intelligence, this important problem can now be tackled via machine learning algorithms that can discover rich and powerful representations for the data at hand. In collaboration, TU Munich and Stanford University plan to devise and implement generative methods that can explain uncertainty and ambiguity in pose predictions. In particular, our aim is to bridge the gap between 6DoF pose estimation either from 2D images/3D point sets and uncertainty quantification through multimodal variational deep methods. Further details can be found here: .

Collaboration:

, , Stanford University

, , , Technical University of Munich

Understanding the effects of artifacts on automated polyp detection and incorporating that knowledge via learning without forgetting

Wed, 01 Jan 2020 00:00:00 +0000

Organizing Committee Member at MICCAI COMPAY 2019

Sun, 13 Oct 2019 00:00:00 +0000

Organizing Committee Member at MICCAI DART 2019

Sun, 13 Oct 2019 00:00:00 +0000

Keynote Speaker: Towards Deep Federated Learning in Healthcare

Thu, 19 Sep 2019 00:00:00 +0000

Method for determining a pose of an object in an environment of the object using multi task learning and control device

Mon, 01 Jul 2019 00:00:00 +0000

Adaptive image-feature learning for disease classification using inductive graph networks

Tue, 01 Jan 2019 00:00:00 +0000

Adversarial Networks for Camera Pose Regression and Refinement

Tue, 01 Jan 2019 00:00:00 +0000

Deep autoencoding models for unsupervised anomaly segmentation in brain MR images

Tue, 01 Jan 2019 00:00:00 +0000

Domain Adaptation and Representation Transfer and Medical Image Learning with Less Labels and Imperfect Data: First MICCAI Workshop, DART 2019, and First International Workshop, MIL3ID 2019, Shenzhen, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 13 and 17, 2019, Proceedings

Tue, 01 Jan 2019 00:00:00 +0000

European Congress on Digital Pathology

Tue, 01 Jan 2019 00:00:00 +0000

Focal loss for artefact detection in medical endoscopy

Tue, 01 Jan 2019 00:00:00 +0000

Generalizing multistain immunohistochemistry tissue segmentation using one-shot color deconvolution deep neural networks

Tue, 01 Jan 2019 00:00:00 +0000

Graph Convolution Based Attention Model for Personalized Disease Prediction

Tue, 01 Jan 2019 00:00:00 +0000

InceptionGCN: receptive field aware graph convolutional network for disease prediction

Tue, 01 Jan 2019 00:00:00 +0000

Investigation of Focal Loss in Deep Learning Models For Femur Fractures Classification

Tue, 01 Jan 2019 00:00:00 +0000

Learn to estimate labels uncertainty for quality assurance

Tue, 01 Jan 2019 00:00:00 +0000

Learn to Segment Organs with a Few Bounding Boxes

Tue, 01 Jan 2019 00:00:00 +0000

Learning interpretable disentangled representations using adversarial vaes

Tue, 01 Jan 2019 00:00:00 +0000

Learning Interpretable Features via Adversarially Robust Optimization

Tue, 01 Jan 2019 00:00:00 +0000

Learning-based x-ray image denoising utilizing model-based image simulations

Tue, 01 Jan 2019 00:00:00 +0000

LYSTO: Benchmark and hackathon for lymphocyte assessment

Tue, 01 Jan 2019 00:00:00 +0000

Reliable assessment of tumor-infiltrating lymphocytes in immunohistochemistry is clinically relevant for prognosis and immunotherapy decision-making but remains time-consuming and subject to inter-observer variability. Within the international LYSTO initiative, I contributed to the development and consolidation of a benchmark initially introduced as a MICCAI 2019 hackathon. The benchmark comprises multi-center, multi-organ IHC datasets with over 170,000 cell-level annotations and formulates lymphocyte assessment as a weakly supervised learning problem. Multiple methods achieved pathologist-level agreement across institutions and organs, and extended analyses incorporating community feedback culminated in a comprehensive peer-reviewed publication in 2024.

Our role

Contributing collaborator with responsibility for methodological analysis, evaluation, and consolidation of benchmark results.

Perspective

LYSTO emphasizes the importance of sustained, community-driven benchmarking that extends beyond time-limited challenges. The experience informed subsequent work on multi-center robustness, evaluation, and trustworthy assessment of medical AI.

Machine Learning and Medical Engineering for Cardiovascular Health and Intravascular Imaging and Computer Assisted Stenting: First International Workshop, MLMECH 2019, and 8th Joint International Workshop, CVII-STENT 2019, Held in Conjunction with MICCAI 2019, Shenzhen, China, October 13, 2019, Proceedings

Tue, 01 Jan 2019 00:00:00 +0000

MRI to CT Translation with GANs

Tue, 01 Jan 2019 00:00:00 +0000

Multi-scale Microaneurysms Segmentation Using Embedding Triplet Loss

Tue, 01 Jan 2019 00:00:00 +0000

Perceptual Embedding Consistency for Seamless Reconstruction of Tilewise Style Transfer

Tue, 01 Jan 2019 00:00:00 +0000

Preliminary results of DSA denoising based on a weighted low-rank approach using an advanced neurovascular replication system

Tue, 01 Jan 2019 00:00:00 +0000

Self-attention equipped graph convolutions for disease prediction

Tue, 01 Jan 2019 00:00:00 +0000

Semi-Supervised Few-Shot Learning with Local and Global Consistency

Tue, 01 Jan 2019 00:00:00 +0000

Staingan: Stain style transfer for digital histological images

Tue, 01 Jan 2019 00:00:00 +0000

Towards an Interactive and Interpretable CAD System to Support Proximal Femur Fracture Classification

Tue, 01 Jan 2019 00:00:00 +0000

Virtualization of tissue staining in digital pathology using an unsupervised deep learning approach

Tue, 01 Jan 2019 00:00:00 +0000

Fusing unsupervised and supervised deep learning for white matter lesion segmentation

Thu, 13 Dec 2018 00:00:00 +0000

Determination of the pose of an x-ray unit relative to an object using a digital model of the object

Mon, 01 Oct 2018 00:00:00 +0000

Invited Talk: Can Deep Learning Models be Trained with Annotations Collected via Crowdsourcing?

Sun, 09 Sep 2018 00:00:00 +0000

StainGAN — Stain Normalization

Wed, 04 Apr 2018 00:00:00 +0000

StainGAN

StainGAN implementation based on Cycle-Consistency Concept

For more information visit .

Structure

Stain-Transfer Model
Pre-processing.
Post-processing.
Evaluation

Datasets

The evaluation was done using the Camelyon16 challenge ( ) consisting of 400 whole-slide images collected in two different labs in Radboud University Medical Center (lab 1) and University Medical Center Utrecht (lab 2). Otsu thresholding was used to remove the background, Afterwards, 40, 000 256 × 256 patches were generated on the x40 magnification level, 30, 000 were used for training and 10, 000 used for validation from lab 1 and 10, 000 patches were generated for testing from lab 2.

Patches can be found here:

Any use of the dataset or anypart of the code should be cited

Citation

If you use this code for your research, please cite our papers.

@inproceedings{shaban2019staingan,
 author = {Shaban, M Tarek and Baur, Christoph and Navab, Nassir and Albarqouni, Shadi},
 booktitle = {2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019)},
 organization = {IEEE},
 pages = {953--956},
 title = {Staingan: Stain style transfer for digital histological images},
 year = {2019}
}

Acknowledgments

Code is inspired by and .

A photon recycling approach to the denoising of ultra-low dose X-ray sequences

Mon, 01 Jan 2018 00:00:00 +0000

Capsule networks against medical imaging data challenges

Mon, 01 Jan 2018 00:00:00 +0000

Deep learning with synthetic data for free water elimination in diffusion MRI

Mon, 01 Jan 2018 00:00:00 +0000

Domain and geometry agnostic CNNs for left atrium segmentation in 3D ultrasound

Mon, 01 Jan 2018 00:00:00 +0000

Fine-Tuning Deep Learning by Crowd Participation

Mon, 01 Jan 2018 00:00:00 +0000

Generating highly realistic images of skin lesions with GANs

Mon, 01 Jan 2018 00:00:00 +0000

Imaging intelligence: AI is transforming medical imaging across the imaging spectrum

Mon, 01 Jan 2018 00:00:00 +0000

Intraoperative stent segmentation in X-ray fluoroscopy for endovascular aortic repair

Mon, 01 Jan 2018 00:00:00 +0000

Intravascular Imaging and Computer Assisted Stenting and Large-scale Annotation of Biomedical Data and Expert Label Synthesis: 7th Joint International Workshop, CVII-STENT 2018 and Third International Workshop, LABELS 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 16, 2018, Proceedings

Mon, 01 Jan 2018 00:00:00 +0000

Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis

Mon, 01 Jan 2018 00:00:00 +0000

LABELS 2018 Preface

Mon, 01 Jan 2018 00:00:00 +0000

Multiple device segmentation for fluoroscopic imaging using multi-task learning

Mon, 01 Jan 2018 00:00:00 +0000

Scene coordinate and correspondence learning for image-based localization

Mon, 01 Jan 2018 00:00:00 +0000

Weakly-supervised localization and classification of proximal femur fractures

Mon, 01 Jan 2018 00:00:00 +0000

When regression meets manifold learning for object recognition and pose estimation

Mon, 01 Jan 2018 00:00:00 +0000

X-ray Depthmaps: Revealing the Hidden Structures

Mon, 01 Jan 2018 00:00:00 +0000

Automatic classification of proximal femur fractures based on attention models

Sun, 01 Jan 2017 00:00:00 +0000

CAD model-based detection and tracking of large objects

Sun, 01 Jan 2017 00:00:00 +0000

Industry project with Siemens AG on CAD model-based detection and tracking of large objects in RGB(D) video (2017–2019).

Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer

Sun, 01 Jan 2017 00:00:00 +0000

Fully Automated Multiple Sclerosis lesion detection on multi-channel subtraction images through an integrated Computer Vision-Machine Learning pipeline

Sun, 01 Jan 2017 00:00:00 +0000

Semi-supervised deep learning for fully convolutional networks

Sun, 01 Jan 2017 00:00:00 +0000

X-ray in-depth decomposition: Revealing the latent structures

Sun, 01 Jan 2017 00:00:00 +0000

X-Ray PoseNet: 6 DoF pose estimation for mobile X-Ray devices

Sun, 01 Jan 2017 00:00:00 +0000

Anomaly Detection

Wed, 20 Apr 2016 00:00:00 +0000

Aggnet: deep learning from crowds for mitosis detection in breast cancer histology images

Fri, 01 Jan 2016 00:00:00 +0000

Cathnets: detection and single-view depth prediction of catheter electrodes

Fri, 01 Jan 2016 00:00:00 +0000

Deep learning and data labeling for medical applications

Fri, 01 Jan 2016 00:00:00 +0000

Deep Learning and Data Labeling for Medical Applications: First International Workshop, LABELS 2016, and Second International Workshop, DLMIA 2016, Held in Conjunction with MICCAI 2016, Athens, Greece, October 21, 2016, Proceedings

Fri, 01 Jan 2016 00:00:00 +0000

Optimisation of X-ray images with respect to image quality and dose

Fri, 01 Jan 2016 00:00:00 +0000

Industry project with Siemens Healthineers on learning-based X-ray image-quality and dose optimisation (2016–2019).

Parsing human skeletons in an operating room

Fri, 01 Jan 2016 00:00:00 +0000

Single-view X-ray depth recovery: toward a novel concept for image-guided interventions

Fri, 01 Jan 2016 00:00:00 +0000

Structure-preserving color normalization and sparse stain separation for histological images

Fri, 01 Jan 2016 00:00:00 +0000

Gradient projection for regularized cryo-electron tomographic reconstruction

Thu, 01 Jan 2015 00:00:00 +0000

Multi-scale Graph-based Guided Filter for De-noising Cryo-Electron Tomographic Data

Thu, 01 Jan 2015 00:00:00 +0000

Structure-preserved color normalization for histological images

Thu, 01 Jan 2015 00:00:00 +0000

Developing MATLAB software for PV and battery sizing for lighting projects in Gaza Strip, Palestine

Tue, 18 Dec 2012 00:00:00 +0000

Developing empirical models for estimating global solar radiation in Gaza Strip, Palestine

Fri, 01 Jan 2010 00:00:00 +0000

Re-evaluation and re-design stand-alone PV solar lighting projects in Gaza Strip, Palestine

Fri, 01 Jan 2010 00:00:00 +0000

Study on the Optimum Tilt Angle and Orientation for Photovoltaic Panels and Feasibility Study of One axis-two positions tracking Solar PV in Palestine

Fri, 01 Jan 2010 00:00:00 +0000

AI for Medical Diagnosis and Prediction

Mon, 01 Jan 0001 00:00:00 +0000

AI in Medicine (AiM): Foundations, Methods, and Critical Perspectives

Mon, 01 Jan 0001 00:00:00 +0000

Artificial Intelligence in Radiology: Applications and Research

Mon, 01 Jan 0001 00:00:00 +0000

Computational Medical Imaging (Lab)

Mon, 01 Jan 0001 00:00:00 +0000

Computational Medical Imaging (Seminar)

Mon, 01 Jan 0001 00:00:00 +0000