Shades of Certainty : Annotation and Classification of Swedish Medical Records
Sumithra Velupillai · 2012
Access to information is fundamental in health care. Today, with electronic documentation possibilities, techniques for automatic extraction of information from written documentation are used daily in many areas. However, in the clinical setting, written documentation is still unattainable for improving health care from many perspectives. For Swedish, research for improving automatic information access from medical records is still scarce. This thesis presents research on Swedish medical records with the overall goal of building intelligent information access tools that can aid health personnel, researchers and other professions in their daily work, and, ultimately, improve health care in general. First, the issue of ethics and identifiable information in medical records is addressed by creating an annotated gold standard corpus for de-identification, and by porting an existing de-identification software to Swedish from English. The aim is to move towards making textual resources that do not risk exposure of an individual patient’s information available to researchers. Results for the ported rule-based system are not encouraging, but the Inter-Annotator Agreement results for the created gold standard are fairly high.