Block Course · Summer & Winter Schools
Summer School on Biomedical Imaging with Deep Learning (BILD) (BILD)
A five-day Arab-German summer school on deep learning for biomedical imaging, organised by the Albarqouni Lab under the SANAD network with the Lebanese American University, the University of Tunis El Manar and Duhok Polytechnic University. Expert lectures, a keynote and daily hands-on sessions take participants from deep-learning fundamentals to segmentation, uncertainty, graph learning, vision-language models and deployable medical AI apps.
About the course
BILD 2025 is the first edition of an Arab-German summer school on deep learning for medical imaging, organised by the Albarqouni Lab at the University Hospital Bonn and the University of Bonn under the umbrella of the Strategic Arab-German Network for Affordable and Democratized AI in Healthcare (SANAD). It is run with partners in the region: the Lebanese American University, the University of Tunis El Manar and Duhok Polytechnic University. Over five intensive days in Tunis, participants follow expert lectures, a keynote and daily hands-on sessions, and work with peers and faculty from Germany, Lebanon, Iraq and Tunisia on building and deploying AI models for real-world healthcare challenges.
Course structure
| Session | Lecture and topics | Lab |
|---|---|---|
| Day 1 | L1 Why AI in medicine (primer) clinical pressure points; promise and evidence; what goes wrong; humans and AI together | you, a model, and the two of you together (blood cells, chest X-ray) |
| L5 Deep learning for classification and detection perceptron to CNN; architectures and transfer; detection, IoU, mAP; class imbalance | filters, small CNNs, rare findings and boxes | |
| Day 2 | L7 Segmentation classic methods; U-Net and nnU-Net; Dice, HD95, losses; few labels and SAM; registration and tracking (rigid to deformable, mutual information, lesion and cell tracking) | outline every organ, then find where it moved (CT pack) |
| L4 Evaluation with clinical metrics (primer) confusion matrix; ROC and PR; prevalence and operating points; confidence intervals | clinical metrics on chest X-rays | |
| Day 3 | L8 Graphs and geometric deep learning graphs in medicine; Laplacian and spectral methods; message passing and GNNs; graph foundation models | relations, not rows |
| L11 Uncertainty quantification aleatoric and epistemic; MC dropout and ensembles; calibration; refer or decide; conformal prediction; uncertainty for graphs and LLMs | calibration, answer sets and shift | |
| Day 4 | L9 Foundation models and multimodal data self-supervision; multimodal fusion and contrastive learning; vision–language models (CLIP); clinical LLMs and agents | foundation models and multimodal data |
| L10 Interpretability and explainability (primer) why XAI; taxonomy; attribution methods; evaluating explanations; explaining GNNs and VLMs | explanations you can test | |
| Day 5 | L12 Bias, fairness and causality (primer) causal diagrams; dataset shift; fairness metrics; mitigation; bias in graphs and foundation models | subgroups, shift and causes (incl. a skin-tone audit of a public dermatology dataset) |
| L13 Trustworthy AI in practice (primer) ethics and the EU AI Act; evidence and reporting; monitoring and drift; federated learning and privacy; carbon footprint | rules, federation, monitoring, footprint |
Toolbox
- T1 Vectors, matrices and gradients: scalars to tensors; dot products and linear maps; distances, Euclidean to geodesic; convexity; gradient, Jacobian, Hessian
- T2 Probability and statistics: z-scores and outliers; distributions, Bernoulli to Rician; CLT and confidence intervals; Bayes and Bayesian updating; MLE, MAP, entropy, KL
- T3 Python toolbox: notebooks and Colab; NumPy on the course CT; pandas; scikit-learn and PyTorch cheat sheets · Lab: arrays, tables, models
Learning objectives
By the end of the school, participants will:
- Understand the fundamentals of deep learning and apply them to medical image classification, segmentation and detection.
- Explain and evaluate segmentation methods and metrics (Dice, Jaccard), and strategies for scarce, weak or noisy annotations.
- Estimate and interpret model uncertainty, and use graph learning and vision-language models (CLIP, MedCLIP) for medical data.
- Develop medical AI with MONAI and perform quality control of models and data.
- Build an end-to-end mobile app prototype for pneumonia detection from chest X-rays.
The school’s teaching modules combine current AI research with resource-aware exercises, such as lightweight networks for portable ultrasound and federated learning on decentralised data, tailored to low-resource settings in the MENA region.
Target audience
Semesters
| Semester | Course | Format | Audience | Lecturers & tutors |
|---|---|---|---|---|
| SS 2025 | Summer School on Biomedical Imaging with Deep Learning (BILD) (BILD) | Block Course | Summer & Winter Schools | Rached Abdelkhalak |