AuMEDa: Automated information and feature extraction from MEDical documents – structured data with artificial intelligence
Partner: University Medical Center Göttingen (UMG)
Project duration: 01/10/2021 – 30/09/2024
Funding: University Medical Center Göttingen (UMG) and University of Göttingen
Methods: Process analysis and Software development
Summary:
Patient information is still exchanged manually in most healthcare institutions in Germany. The current process is not only time-consuming for medical staff but also error-prone due to repetitive manual typing of text into multiple systems. AuMEDa aims to create a platform for automated information and feature extraction from paper-based medical documents by using different technology stacks such as artificial intelligence. To meet the high requirements for information in healthcare, the system will include a validation interface for medical staff in order to control extracted information. Through reinforcement learning, feedback of the experts will be used to improve the system continuously. The extracted information will be saved in a data lake, which can then be utilized for patient care and research purposes.
The DHRG will be responsible for the development of the platform.