Data reduction for continuum of care: An exploratory study using the predicate-argument structure to pre-process radiology sentences for measurement of semantic similarity

Abstract

In the clinical setting, continuum of care depends on integrated information services to assure a smooth progression for patient centered care, and these integrated information services must understand past events and personal circumstances to make care relevant. Clinicians face a problem that the amount of information produced in disparate electronic clinical notes is increasing to levels incapable of being processed by humans. Clinicians need a function in information services that can reduce the free text data to a message useful at time of care. Information extraction (IE) is a sub-field of natural language processing with the goal of data reduction of unstructured free text. Pertinent to IE is an annotated corpus that frames how IE methods should create a logical expression necessary for processing meaning of text. This study explores and reports on the requirements to using the predicate-argument statement (PAS) as the framework. A convenient sample from a prior study with ten synsets of 100 unique sentences from radiology reports deemed by domain experts to mean the same thing will be the text from which PAS structures are formed. Through content analysis of pattern recognition, findings show PAS is a feasible framework to structure sentences for semantic similarity measurement. © 2013 Springer-Verlag Berlin Heidelberg.