Workshop: Sharing Healthcare Records for NLP

  • Radboud University: Djoerd Hiemstra (Institute for Computing and Information Sciences)
  • External Partners: Nedap Healthcare

With the strong adoption of electronic health records (EHRs), large quantities of unstructured medical patient data become available. These data offer significant opportunities to advance medical research and to improve healthcare related services. Natural language processing (NLP) techniques are commonly used to process, analyze and understand medical documents on a large scale. However, most NLP techniques were developed on English medical data and are not applicable to Dutch data. The lack of openly available Dutch data hampers the progress that researchers and practitioners make in this domain. The Dutch community of researchers and healthcare practitioners requires sharable and public resources to advance text and data-mining techniques for healthcare data.

The goal of this work is to bring together researchers to exchange ideas to share healthcare data sets. With the support of the innovation voucher, we organize an attractive workshop on this topic and start a discussion on how to realise publicly available medical datasets. Possible topics include:

  • Ethical and legal concerns around the sharing of privacy sensitive data,
  • Techniques and resources to anonymize medical data,
  • Common and best practices when working with privacy sensitive data.

This workshop will be jointly organized with our private partner Nedap Healthcare. Nedap Healthcare provides software to Dutch healthcare providers working in elderly care, mental health care, and disabled care. Their software suite, Nedap Ons, supports the full care process, ranging from planning and giving care to the financial administration and communication between care practitioners. One of the primary AI challenges faced by Nedap developers is adapting solutions for English data to Dutch.

Stay tuned for more information on the workshop on this website!