Automatic meaning discrimination by a machine is a complex task that is critical to accessing information encoded in clinical texts 6, 7. The process of understanding the precise meaning of a given acronym or abbreviation in texts is one of several key functions of automated medical natural language processing (NLP) systems 4 and is a special case of word sense disambiguation (WSD) 5. Several reasons for this ongoing growth include adoption of electronic health record (EHR) systems with increased volume of electronic clinical notes accompanied by the wide usage of acronyms and abbreviations 2, the time-constrained nature of clinical medicine encouraging the use of shortened word forms, and a longstanding tradition of commonly using acronyms and abbreviations in clinical documentation 1. These findings support future work in clinical acronym and abbreviation WSD and require validation with other clinical texts.Īcronyms and abbreviations within clinical texts are widespread, and their use continues to increase 1 – 3. To achieve an accuracy of over 90%, the minimum required training sample size was approximately 125 samples for SVM classifiers with inverted cross-validation. Our analysis of window size and orientation showed best performance using a larger left-sided and smaller right-sided window. In this paper, fifty clinical acronyms and abbreviations with 500 samples each were studied using supervised machine-learning techniques (Support Vector Machines (SVM), Naïve Bayes (NB), and Decision Trees (DT)) to optimize the window size and orientation and determine the minimum training sample size needed for optimal performance. Automated acronym sense disambiguation (WSD), a task of assigning the context-appropriate sense to ambiguous clinical acronyms and abbreviations, represents an active problem for medical natural language processing (NLP) systems. Acronyms and abbreviations within electronic clinical texts are widespread and often associated with multiple senses.
0 kommentar(er)
