Meet our Project Partner: University of Birmingham

UoB role in INSAFEDARE

The University of Birmingham leads key technical and methodological contributions within INSAFEDARE, focusing on standards, data assurance, and synthetic data generation.

We are responsible for reviewing and formally modelling existing standards and guidance on data safety and quality, developing a standards ontology to enable interoperability across regulatory frameworks. We also investigate controls and evidence for data-related safety assurance, identifying data failure modes and defining statistical and methodological approaches to mitigate risks. Finally, we lead the development of advanced artificial intelligence methods for generating synthetic medical imaging data, with a focus on brain MRI imaging.

These contributions strengthen the project’s toolkit by providing robust assurance frameworks, validated methodologies, and innovative synthetic datasets to support safe and effective data-driven validation of medical devices.

Key progress in the last 3 months

• Assisted EFMI on the completion and submission of deliverable D3.1.

• Lead and submitted deliverable D3.2, with Warwick as the task leader.

• Task leader for deliverable D2.2. Reviewed standards and guidance on data quality and safety in healthcare. The deliverable has been sent out to the consortium and received positive feedbacks.

• Task leader for deliverable D2.3. Developed a standards ontology to model the regulatory frameworks of data quality and safety in a machine-readable format. The deliverable is planned to be sent out to the consortium soon.

• Task and deliverable leader for deliverable D4.3. Trained generative AI models for the generation of synthetic medical imaging data and evaluated the synthetic data with a proposed evaluation framework to ensure their fidelity, utility, diversity, and privacy. The deliverable is planned to be sent out to the consortium soon.

Publications and manuscripts (published or in preparation)

Published:

• Abbasi, Saadullah Farooq, et al. "Deep learning-based synthetic skin lesion image classification." 34th Medical Informatics Europe Conference, MIE 2024. IOS Press, 2024.

• Abbasi, Saadullah Farooq, et al. "Preliminary Results on Improved Synthetic Image Generation for Melanoma Skin Cancer." International Conference on Informatics, Management, and Technology in Healthcare 2024. IOS Press, 2025.

• Abbasi, Saadullah Farooq, et al. "A Novel and Secure 3D Colour Medical Image Encryption Technique Using 3D Hyperchaotic Map, S-box and Discrete Wavelet Transform." Studies in health technology and informatics 328 (2025): 268-272.

• Ding, Xuefei, et al. "Recent Advances in Generative Models for Synthetic Brain MRI Image Generation." Global Healthcare Transformation in the Era of Artificial Intelligence and Informatics (2025): 51-55.

• Noor, N., Bilal, M., Abbasi, S.F., Pournik, O. and Arvanitis, T.N., 2025. A novel transformer-based approach for cardiovascular disease detection. Frontiers in digital health, 7, p.1548448.

In preparation:

• Conference and/or journal publications for the ontology modelling of standards on data safety and quality in healthcare (D2.2, D2.3).

• Conference and/or journal publications for systematic reviews on privacy-preserved synthetic medical images generated by AI (D4.3).

• Conference and/or journal publications for systematic reviews on the evaluation frameworks for privacy-preserved synthetic medical images (D4.3).