Pestian et al BioNLP 2007
Citation
Pestian et al. 2007. A Shared Task Involving Multi-label Classification of Clinical Free Text. In Proceedings of the BioNLP 2007, 97-104.
Online version
Summary
The paper presents a MEDical Information Extraction (MedIE) system, which extracts patient information from free-text clinical records.
They divided their extraction job into three tasks below.
- extraction of medical terms
- relation extraction
- extraction of associated medical concepts
- e.g. Blood pressure & 144/90 in the sentence, "Blood pressure is 144/90"
- text classification
- e.g. a patient can be classified as a former smoker, a current smoker, or a non-smoker
Their approaches are:
- An ontology-based approach for extracting medical terms of interest
- they used Unified Medical Language System (UMLS)
- About terms that are not defined in UMLS, they predicted categories of some terms using sentence structures.
- A graph-based approach which uses the parsing result of link-grammar parser for relation-extraction
- They included the processing of negation.
- When the parser fails, they used a pattern-based approach.
- Because the parser did not process multi-word terms, they replaced the terms with placeholders.
- an NLP-based feature extraction method coupled with an ID3-based decision tree for text classification
This approach was fairly successful mostly showing over 80% of precision and recall. However, the system was tested on the data written by only a clinician, which means that the style of free-text records was consistent. Nevertheless, the research is worth in that they applied various IE techniques to the free-text clinical records, explain about the problems they encountered.
Related papers
The widely cited Pang et al EMNLP 2002 paper was influenced by this paper - but considers supervised learning techniques. The choice of movie reviews as the domain was suggested by the (relatively) poor performance of Turney's method on movies.
An interesting follow-up paper is Turney and Littman, TOIS 2003 which focuses on evaluation of the technique of using PMI for predicting the semantic orientation of words.
