Text mining and information analysis of health documents

作者:Suominen; Hanna 刊名:Artificial Intelligence in Medicine 上传者:李远方



Artificial Intelligence in Medicine 61 (2014) 127–130 Contents lists available at ScienceDirect Artificial Intelligence in Medicine j o ur na l ho mepage: www.elsevier.com/locate/aiim Guest Editorial Text mining and information analysis of health documents 1. Introduction Documenting every event in healthcare, as required by law in many countries, takes a lot of healthcare professionals’ time. With electronic health (eHealth) documents being increasingly common, healthcare professionals type approximately forty per cent of the documents as free-form text and the remaining sixty per cent is either automatically recorded or typed structured-information [1]. For example, one patient’s intensive care documents alone can con- tain over sixty typed pages of free-form text [2], supplemented each day by over 1500 structured items, on average [3]. Free-form text as an entry type is essential to release healthcare professionals’ time from documentation for other tasks. With fully structured or centralized eHealth documents, entering information can take nearly sixty per cent of their working time, whilst with electronic records that allow free text entry at the point of care, it typically takes only a few minutes per patient [4–6]. Even more time could be released through human/natural language technologies (HLTs a.k.a. computational linguistics, human/natural language processing, text analytics, text (data) min- ing, or text processing (TM)) to support healthcare professionals in genera