The healthcare research programme has a strategic focus to support the assurance-based adoption of AI in healthcare - building on knowledge we gained through our whitepaper How do I turn this on? What to consider when adopting AI-based tools into clinical practice.
DNV’s Healthcare Research group aims to develop new roles in assurance, cybersecurity, quality, privacy, regulatory and digitalisation by participation in public-private projects with academic and industry partners. Our current project portfolio includes projects on federated health networks, synthetic data, healthcare interoperability and data portability and health technology assessment (HTA) of digital health technologies, funded by IHI, Nordic Innovation , Horizon Europe and an NFR funded PhD.
SYNTHIA aims to create validated, reliable tools and methods for generation of synthetic data that mimics real patient data to address six diseases: lung and breast cancer, multiple myeloma and diffuse large B-cell lymphoma, Alzheimer’s disease and type-2 diabetes. DNV will ensure trust through development of a framework for assurance of different synthetic data modalities and will investigate the implications of utilising synthetic data within applications that involve regulatory bodies in healthcare.
FederatedHealth aims to harness the potential of unstructured data in Electronic Health Record (EHR) systems by developing a federated health data network in the Nordics, using distributed machine learning to ensure data privacy while processing multilingual clinical text from several languages. DNV is leading the identification and analysis of barriers for implementing the federated learning approach and mapping solutions to overcome these. DNV is also assessing the data and model security chosen for the federated learning infrastructure.
EDiHTA aims to create the first flexible, inclusive, validated and ready-for-use European health technology assessment (HTA) framework allowing the systematic assessment of different Digital Health Technologies to inform healthcare decision-makers of the rationale for their implementation. DNV will contribute to the creation of the framework with cybersecurity and regulatory expertise. DNV will also lead the development of the digital solution for operationalization of the EDiHTA framework in the EU.
xShare aims to enable the European population to share their health data in EEHRxF with a click-of-a-button through co-design of a standards and policy hub and creation of the xShare industry label. DNV will contribute to the development of recommendations for European Electronic Health Record eXchange Format (EEHRxF) testing and assurance and capacity building on eID, security and privacy in health.
In an NFR funded PhD project, DNV with partners Oslo University Hospital, the Cancer Registry of Norway and the University of Oslo is exploring how the use of synthetic data for AI development and validation can be assessed to ensure safe implementation of AI in healthcare. As healthcare data, with its inherently sensitive nature, is often challenging to share and process, synthetic data is increasingly seen as a practical way to speed up the development process while protecting patient privacy. The project will also investigate reidentification risk and their legal consequences for different use case scenarios.
Can I trust my fake data – A comprehensive quality assessment framework for synthetic tabular data in healthcare
This paper presents a conceptual framework for quality assurance of synthetic data for AI applications in healthcare
Adoption of AI in healthcare
In this white paper we present considerations to help facilitate the safe and widespread adoption of AI-based tools in healthcare.
A Systematic Literature Review of User Trust in AI-Enabled Systems: An HCI Perspective
This review aims to provide an overview of the user trust definitions, influencing factors, and measurement methods from 23 empirical studies to gather insight for future technical and design strategies, research, and initiatives to calibrate the user-AI relationship.
Challenges, Needs and Opportunities In Federated Health Data Networks
Federated health data networks (FHDNs) have emerged as an attractive solution to ‘freeing’ health data.