Synthetic EHR study

Creating an Open, Synthetic Medical Record Dataset in Swedish for AI Development in Healthcare

The project seeks to address the scarcity of accessible data for AI development in Swedish healthcare by generating a collection of open, synthetic medical records in Swedish. These records will represent fictional patients and scenarios, but their language and structure will be modeled after real patient records to ensure linguistic realism. Advanced algorithms and rigorous evaluation methods will be employed to guarantee that the records are nearly indistinguishable from authentic ones while fully safeguarding patient privacy.

With a target of producing 2,000 high-quality synthetic medical records, the project will establish a publicly accessible repository for researchers and developers. This resource aims to accelerate the development and deployment of AI models capable of processing and understanding Swedish medical records, ultimately contributing to more effective AI applications in healthcare.

The Team Behind It

• David Sundemo
• Jonathan Widén
• Birger Moell
• Jens Lundström
• Max Petzold

Names listed alphabetically

Project Status

The project is in the planning phase. The next milestone will be the ethics approval application.

Funding

This project is supported by Forte.