This talk will present briefly the field of BioMedical Natural Language Processing where the first essential task is to identify automatically the entities (terms) and the indicators’ values in the free text.
Recently the aim is to analyse outpatient records submitted to the Bulgarian National Insurance Fund. Relying on the relatively structured hypertext format of the outpatient records, our tools search only in the sections Anamnesis and Patient status. Two extractors are integrated in the software environment behind the Diabetic register: for extraction of glycated hemoglobin and values of blood pressure. This enables to make more detailed explorations of the diabetes development in Bulgaria.
Dr. Galia Angelova, PhD, DSc is Professor in Computer Science at the Institute of Information and Communication Technologies, Bulgarian Academy of Sciences, Sofia. Her research fields are Computational Linguistics and Artificial Intelligence. Since more than 10 years she works in the area of automatic analysis of clinical narratives in Bulgarian language. Prof. Angelova coordinated several projects with national or European funding where software tools for Information Extraction from Bulgarian patient records were developed and gradually improved (one tool that discovers drugs taken by the patient at the moment of hospitalisation is presented in the paper "Contextualization in automatic extraction of drugs from Hospital Patient Records"; this paper received the Rolf Hansen Memorial Award for best scientific paper on Electronic Health Records at EFMI 2011). Tools for automatic text analysis are integrated in the Bulgarian Diabetic register and perform extraction from a large-scale repository of millions of outpatient records.