Healthcare information systems collect massive amounts of textual and numeric information about patients, visits, prescriptions, physician notes and more. The information encapsulated within electronic clinical records could lead to improved healthcare quality, promotion of clinical and research initiatives, fewer medical errors and lower costs. However, the documents that comprise the health record vary in complexity, length and use of technical vocabulary. This makes knowledge discovery complex. Commercial text mining tools provide a unique opportunity to extract critical information from textual data archives. In this paper, we share our experience of a collaborative research project to develop predictive models by text mining electronic clinical records. We provide an overview of the text mining process, examples of existing studies, experiences of our collaborative project and future opportunities.
[1]
Brian Hazlehurst,et al.
Detecting Possible Vaccination Reactions in Clinical Notes
,
2005,
AMIA.
[2]
John F. Hurdle,et al.
Automated identification of adverse events related to central venous catheters
,
2007,
J. Biomed. Informatics.
[3]
Carol Friedman,et al.
Research Paper: Human and Automated Coding of Rehabilitation Discharge Summaries According to the International Classification of Functioning, Disability, and Health
,
2006,
J. Am. Medical Informatics Assoc..