On 5 December, the project hosted a Community of Practice (CoP) webinar focused on recent advances in synthetic data generation and evaluation for medical applications.
The session brought together experts working on state-of-the-art approaches for data augmentation, privacy preservation, and model robustness assessment, covering both tabular clinical data and medical imaging use cases. Real-world applications discussed included cardiac decompensation prediction, pneumonia detection, Parkinson’s disease progression, epilepsy surgery, and synthetic retinal imaging.
The webinar was moderated by Mariana Oliveira and featured the following presentations:
- Pedro Matias & Maria Russo – Synthetic Clinical Data Augmentation for Predicting Cardiac Decompensation and Pulmonary Exacerbation
- Ivo Façoco – Adapting Stable Diffusion Models for Domain-Specific Medical Imaging: Synthetic Retinal Fundus Image Generation
- Aníbal Silva – Generation and Evaluation of Synthetic Tabular Data for Advanced Parkinson Disease Progression
- Inês Gomes – Using Ambiguous Synthetic Data to Evaluate Model Overconfidence in Pneumonia Detection
- Inês Silveira & Luís Silva – Generation and Evaluation of Synthetic Intraoperative Electrocorticography (ioECoG) for Epilepsy Surgery
The event fostered valuable discussion on how synthetic data can support trustworthy AI development in healthcare while addressing data scarcity and privacy challenges.
