Detecting information in electronic health record texts

126 views
Skip to first unread message

Bas van Stein

unread,
Mar 21, 2022, 10:49:15 AMMar 21
to LIACS thesis projects
We have various projects available for structuring information appearing in free text in electronic health records related to

- anonymisation: recognising and masking sensitive patient information (such as patient name) in text data
- negation: recognition of negated symptoms/remarks (e.g. X individual did not display any reaction to medication Y, X is unwell, X is not aggressive, etc.)
- treatment or disease prognosis

We would like to investigate either knowledge-based or deep learning approaches and/or combine and evaluate existing approaches on real world data in Dutch.

The data available include: general practitioners patient notes, EHR from mental health centres and child psychiatric data.

The projects will be in collaboration with LUMC.

Supervisors: 
dr Kalliopi Zervanou & Prof. Marco Spruit    

For more information please do not hesitate to contact me at: k.zer...@liacs.leidenuniv.nl 

Reply all
Reply to author
Forward
0 new messages