Infrastructure

The Department of Computer and Systems Sciences has several infrastructures for research. Among other things, we have databases, tools and lab environments.

This page is under construction.

 

Swedish Health Record Research Bank (or “Health Bank” for short) is a unique research infrastructure containing large sets of electronic patient records. For example, the Stockholm EPR (Electronic Patient Record) Corpus. The corpus contains data from over 512 clinical units – more than two million patients – at Karolinska University Hospital encompassing the years 2006–2014.

Stockholm EPR Corpus stems from the TakeCare electronic patient records system that is used at the Karolinska University Hospital. All patient records are deidentified. This big data corpus contains both structured information and unstructured information. The structured information contains a serial number for each patient, age, gender, ICD-10 diagnosis codes, drugs but also lab and blood values as well as admission and discharge time and date. The unstructured data contains text written under different headings. The whole corpus contains over 3 227 million tokens.

Health Bank is used and has been used in a number of research projects carried out by the Clinical Text Mining Group. The research is approved by the Regional Ethical Review Board in Stockholm (Regionala Etikprövningsnämnden i Stockholm) and the Swedish Ethical Review Authority (Etikprövningsmyndigheten) under various research plans.
Clinical Text Mining Group

We have developed a number of clinical text mining tools based on the Stockholm EPR Corpus.
Clinical text mining tools

Contact: Hercules Dalianis, Professor, Director Health Bank

For reference to the Health Bank, please use: Dalianis, H., A. Henriksson, M. Kvist, S. Velupillai and R. Weegar. 2015. HEALTH BANK – A Workbench for Data Science Applications in Healthcare (pdf). Proceedings of the CAiSE-2015 Industry Track co-located with 27th Conference on Advanced Information Systems Engineering (CAiSE 2015), J. Krogstie, G. Juell-Skielse and V. Kabilan, (Eds.), Stockholm, Sweden, June 11, 2015, CEUR, Vol-1381, pp 1-18, urn:nbn:de:0074-1381-0.

 

 

On this page