For over 10 years, Electronic Health Records (EHRs) have been a comprehensive source of clinical information for millions of patients in most developed countries. Leveraging the knowledge within these records can decisively aid in accelerating medical research and improving the quality of care, which is essential. Currently, there are already some solutions that require first normalizing and classifying data, and then applying techniques for detecting Protected Health Information (PHI) in clinical texts. However, these solutions are not entirely scalable and often require manual intervention in the first step, resulting in bottlenecks that prevent the global EHR document anonymization process from being fully automated.

This project focuses on the first step of the process so that a reliable system can be developed to provide the necessary scalability for the company’s products while also increasing the quality and anonymity of input data. This will multiply the number of EHR documents that our current products can process and add more value to the company’s customer base, which will be an enormous benefit to the scientific community as a whole.