Privacy compliant health data as a service for AI development

人工智能（AI）使医疗保健领域的数据驱动创新成为可能。人工智能系统可以快速而详细地处理大量数据，作为预防性医疗保健和临床决策的工具，它显示出了希望。然而，分布式存储和对健康数据的有限访问构成了创新的障碍，因为开发值得信赖的人工智能系统需要大型数据集进行训练和验证。此外，匿名数据集的可用性将通过支持卫生技术评估和教育来增加人工智能工具的采用。安全、隐私合规的数据利用是释放人工智能和数据分析全部潜力的关键。在这个建议中，我们将推进目前的国家的最先进的数据合成方法对一个更普遍的方法合成数据生成。我们还将开发用于测试和验证的度量标准，以及能够在不访问真实世界数据的情况下（通过多方计算）生成合成数据的协议。我们的目标是提供：1）改进隐私保护数据合成的方法和技术管道，包括不同的数据格式，如EHR和医学图像，2）易于使用和可配置的数据服务，使AI开发人员能够通过多方计算访问更大的分散去识别数据池，3）按需提供匿名数据或从一个（临时）存储库，4）建立数据市场-促进数据共享和货币化，包括5）将数据市场和数据服务生态系统整合为欧洲健康数据空间中的X-European健康数据中心，以及6）将结果与现实世界的用例相结合，重点关注高影响疾病，特别是癌症类型。

Artificial intelligence (AI) enables data-driven innovations in health care. AI systems, which process vast amounts of data quickly and in detail, show promise both as a tool for preventive health care and clinical decision-making. However, the distributed storage and limited access to health data form a barrier to innovation, as developing trustworthy AI systems requires large datasets for training and validation. Furthermore, the availability of anonymous datasets would increase the adoption of AI-powered tools by supporting health technology assessments and education. Secure, privacy compliant data utilization is key for unlocking the full potential of AI and data analytics. In this proposal, we will advance the current state-of-the-art data synthesis methods towards a more generalized approach of synthetic data generation. We will also develop metricsfor testing and validation, as well as protocolsthat enable synthetic data generation without access to real-world data (through multi-party computation). We aim to provide: 1) Improved methods and technical pipelines for privacy-preserving data synthesis including different data formats such as EHRs and medical images, 2) Easy to use and configurable data servicesto enable AI developers’ accessto larger pools of decentralized de-identified data through multi-party computing, 3) Provide anonymous data on demand or from a (temporary) repository, 4) Establish a Data Market – facilitating data sharing and monetization incl. incentives-based provision of data to the services, 5) Integrate the data market and the data service ecosystem as a X-European health data hub in the European Health Data Space, and 6) Validate the results with real-world use-cases focusing on high impact diseases, cancer types in particular.