Mainz 2026 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
Q: Fachverband Quantenoptik und Photonik
Q 67: Poster – Quantum Information
Q 67.24: Poster
Thursday, March 5, 2026, 17:00–19:00, Philo 2. OG
Synthetic Data Generation for Healthcare Prediction — •Paula Manso Zorilla1, Yannick Werner1,2, Hamraz Javaheri1,2, Gregor Alexander Stavrou3, Omid Ghamarnejad3, Paul Lukowicz1,2, Maximilian Kiefer-Emmanouilidis1,2, and Vitor Fortes Rey1,2 — 1DFKI Kaiserslautern — 2RPTU Kaiserslautern-Landau — 3Department of General, Visceral and Oncological Surgery, Klinikum Saarbrücken
Advancing machine learning in healthcare is often hindered by limited, imbalanced, and privacy-sensitive clinical datasets. To address these constraints, we investigate physics-inspired synthetic data generation using GANs, large language models, and a Quantum Circuit Born Machine trained with maximum mean discrepancy loss. We assess the resulting datasets in terms of fidelity, privacy preservation, and practical utility by training classical and quantum classifiers while evaluating performance exclusively on real patient data. Our findings show that synthetic clinical data can improve robustness and predictive capability in low-data settings. This demonstrates how quantum and classical generative models can help unlock reliable, privacy-preserving AI for real-world healthcare applications.
Keywords: Quantum Machine Learning; Quantum for Health; Quantum Generative Models