Automated misspelling detection and correction in clinical free-text records

Correct record is important in all fields especially in the medical field. For correct medication, safety of patient and correct documentation accurate medical records are important. There is possibility that medical record can contain some misspelled words that can lead towards the wrong interpretation and wrong treatment of patients. So, it is necessary that the medical record must be correct for proper treatment and care of patients. Part of speech tagging is à mechanism that reads the text in documents and separates the different part of speech. Regular expression is a mechanism that is used to match the specific pattern of words in documents. A Methodology/spellchecker is proposed for automated misspelling detection and correction in medical free-text records for correction of misspelled terms. Proposed methodology is based on part of speech tagging and regular expression and is applied on all types of medical documents such as allergy entries, free-text medication order and clinical notes. Dictionary lookup is used for detection of misspelled words. Suffix and prefix based suggestion list is used for misspelled correction. Incorrect spellings from different medical documents are detected and corrected. A significant percentage of misspelling detection and correction is achieved.

[1]  Anabel M Scaranelo,et al.  Error rates in breast imaging reports: comparison of automatic speech recognition and dictation transcription. , 2011, AJR. American journal of roentgenology.

[2]  Lennart Nordenfelt.  Identification and Classification of Diseases: Fundamental Problems in Medical Ontology and Epistemology , 2013 .

[3]  Li Zhou,et al.  Automated misspelling detection and correction in clinical free-text records , 2015, J. Biomed. Informatics.

[4]  Eric Atwell,et al.  Development of tag sets for part-of-speech tagging , 2008 .

[5]  Fred J. Damerau,et al.  A technique for computer detection and correction of spelling errors , 1964, CACM.

[6]  Haruhiko Kaiya,et al.  Ontology based requirements analysis: lightweight semantic processing approach , 2005, Fifth International Conference on Quality Software (QSIC'05).

[7]  B L Lambert,et al.  Predicting look-alike and sound-alike medication errors. , 1997, American journal of health-system pharmacy : AJHP : official journal of the American Society of Health-System Pharmacists.

[8]  Priyanka More,et al.  Generating UML Diagrams from Natural Language Specifications , 2012 .

[9]  Usman Qamar,et al.  Identification and Correction of Misspelled Drugs Names in Electronic Medical Records (EMR) , 2016, ICEIS.

[10]  D. Karthika Renuka,et al.  Email classification for Spam Detection using Word Stemming , 2010 .

[11]  Walter E. Haefeli,et al.  Misspellings in drug information system queries: Characteristics of drug name spelling errors and strategies for their prevention , 2010, Int. J. Medical Informatics.

[12]  Muhammad Ali Babar,et al.  An Automated Tool for Generating UML Models from Natural Language Requirements , 2009, 2009 IEEE/ACM International Conference on Automated Software Engineering.

[13]  Elizabeth D. Liddy,et al.  An NLP Approach for Improving Access to Statistical Information for the Masses , 2001 .

[14]  R. Platt,et al.  Using automated medical records for rapid identification of illness syndromes (syndromic surveillance): the example of lower respiratory infection , 2001, BMC public health.

[15]  Karen Kukich,et al.  Techniques for automatically correcting words in text , 1992, CSUR.