Clinical Text Mining : : Secondary Use of Electronic Patient Records.

• Author: Dalianis, Hercules.
• Place / Publishing House: Cham : : Springer International Publishing AG,, 2018.; Ã2018.
• Year of Publication: 2018
• Edition: 1st ed.
• Language: English
• Physical Description: 1 online resource (192 pages)

Table of Contents

Intro -- Preface -- Objectives -- Organisation of Book Chapters -- Intended Readers -- Limitations -- Book Project During Sabbatical Stay in Sydney -- Aims -- Acknowledgements -- Contents -- 1 Introduction -- 1.1 Early Work and Review Articles -- 2 The History of the Patient Record and the Paper Record -- 2.1 The Egyptians and the Greeks -- 2.2 The Arabs -- 2.3 The Swedes -- 2.4 The Paper Based Patient Record -- 2.5 Greek and Latin Used in the Patient Record -- 2.6 Summary of the History of the Patient Record and the Paper Record -- 3 User Needs: Clinicians, Clinical Researchers and Hospital Management -- 3.1 Reading and Retrieving Efficiency of Patient Records -- 3.2 Natural Language Processing on Clinical Text -- 3.3 Electronic Patient Record System -- 3.4 Different User Groups -- 3.5 Summary -- 4 Characteristics of Patient Records and Clinical Corpora -- 4.1 Patient Records -- 4.2 Pathology Reports -- 4.3 Spelling Errors in Clinical Text -- 4.4 Abbreviations -- 4.5 Acronyms -- 4.6 Assertions -- 4.6.1 Negations -- 4.6.2 Speculation and Factuality -- Levels of Certainty -- Negation and Speculations in Other Languages, Such as Chinese -- 4.7 Clinical Corpora Available -- 4.7.1 English Clinical Corpora Available -- 4.7.2 Swedish Clinical Corpora -- 4.7.3 Clinical Corpora in Other Languages than Swedish -- 4.8 Summary -- 5 Medical Classifications and Terminologies -- 5.1 International Statistical Classification of Diseases and Related Health Problems (ICD) -- 5.1.1 International Classification of Diseases for Oncology (ICD-O-3) -- 5.2 Systematized Nomenclature of Medicine: Clinical Terms (SNOMED CT) -- 5.3 Medical Subject Headings (MeSH) -- 5.4 Unified Medical Language Systems (UMLS) -- 5.5 Anatomical Therapeutic Chemical Classification (ATC) -- 5.6 Different Standards for Interoperability -- 5.6.1 Health Level 7 (HL7).

Fast Healthcare Interoperability Resources (FHIR) -- 5.6.2 OpenEHR -- 5.6.3 Mapping and Expanding Terminologies -- 5.7 Summary of Medical Classifications and Terminologies -- 6 Evaluation Metrics and Evaluation -- 6.1 Qualitative and Quantitative Evaluation -- 6.2 The Cranfield Paradigm -- 6.3 Metrics -- 6.4 Annotation -- 6.5 Inter-Annotator Agreement (IAA) -- 6.6 Confidence and Statistical Significance Testing -- 6.7 Annotation Tools -- 6.8 Gold Standard -- 6.9 Summary of Evaluation Metrics and Annotation -- 7 Basic Building Blocks for Clinical Text Processing -- 7.1 Definitions -- 7.2 Segmentation and Tokenisation -- 7.3 Morphological Processing -- 7.3.1 Lemmatisation -- 7.3.2 Stemming -- 7.3.3 Compound Splitting (Decompounding) -- 7.3.4 Abbreviation Detection and Expansion -- A Machine Learning Approach for Abbreviation Detection -- 7.3.5 Spell Checking and Spelling Error Correction -- Spell Checking of Clinical Text -- Open Source Spell Checkers -- Search Engines and Spell Checking -- 7.3.6 Part-of-Speech Tagging (POS Tagging) -- 7.4 Syntactical Analysis -- 7.4.1 Shallow Parsing (Chunking) -- 7.4.2 Grammar Tools -- 7.5 Semantic Analysis and Concept Extraction -- 7.5.1 Named Entity Recognition -- Machine Learning for Named Entity Recognition -- 7.5.2 Negation Detection -- Negation Detection Systems -- Negation Trigger Lists -- NegEx for Swedish -- NegEx for French, Spanish and German -- Machine Learning Approaches for Negation Detection -- 7.5.3 Factuality Detection -- 7.5.4 Relative Processing (Family History) -- 7.5.5 Temporal Processing -- TimeML and TIMEX3 -- HeidelTime -- i2b2 Temporal Relations Challenge -- Temporal Processing for Swedish Clinical Text -- Temporal Processing for French Clinical Text -- Temporal Processing for Portuguese Clinical Text -- 7.5.6 Relation Extraction -- 2010 i2b2/VA Challenge Relation Classification Task.

Other Approaches for Relation Extraction -- 7.5.7 Anaphora Resolution -- i2b2 Challenge in Coreference Resolution for Electronic Medical Records -- 7.6 Summary of Basic Building Blocks for Clinical Text Processing -- 8 Computational Methods for Text Analysis and Text Classification -- 8.1 Rule-Based Methods -- 8.1.1 Regular Expressions -- 8.2 Machine Learning-Based Methods -- 8.2.1 Features and Feature Selection -- Term Frequency-Inverse Document Frequency, tf-idf -- Vector Space Model -- 8.2.2 Active Learning -- 8.2.3 Pre-Annotation with Revision or Machine Assisted Annotation -- 8.2.4 Clustering -- 8.2.5 Topic Modelling -- 8.2.6 Distributional Semantics -- 8.2.7 Association Rules -- 8.3 Explaining and Understanding the Results Produced -- 8.4 Computational Linguistic Modules for Clinical Text Processing -- 8.5 NLP Tools: UIMA, GATE, NLTK etc -- 8.6 Summary of Computational Methods for Text Analysis and Text Classification -- 9 Ethics and Privacy of Patient Records for Clinical Text Mining Research -- 9.1 Ethical Permission -- 9.2 Social Security Number -- 9.3 Safe Storage -- 9.4 Automatic De-Identification of Patient Records -- 9.4.1 Density of PHI in Electronic Patient Record Text -- 9.4.2 Pseudonymisation of Electronic Patient Records -- 9.4.3 Re-Identification and Privacy -- Black Box Approach -- 9.5 Summary of Ethics and Privacy of Patient Records for Clinical Text Mining Research -- 10 Applications of Clinical Text Mining -- 10.1 Detection and Prediction of Healthcare Associated Infections (HAIs) -- 10.1.1 Healthcare Associated Infections (HAIs) -- 10.1.2 Detecting and Predicting HAI -- 10.1.3 Commercial HAI Surveillance Systems and Systems in Practical Use -- 10.2 Detection of Adverse Drug Events (ADEs) -- 10.2.1 Adverse Drug Events (ADEs) -- 10.2.2 Resources for Adverse Drug Event Detection -- 10.2.3 Passive Surveillance of ADEs.

10.2.4 Active Surveillance of ADEs -- 10.2.5 Approaches for ADE Detection -- An Approach for Swedish Clinical Text -- An Approach for Spanish Clinical Text -- A Joint Approach for Spanish and Swedish Clinical Text -- 10.3 Suicide Prevention by Mining Electronic Patient Records -- 10.4 Mining Pathology Reports for Diagnostic Tests -- 10.4.1 The Case of the Cancer Registry of Norway -- 10.4.2 The Medical Text Extraction (Medtex) System -- 10.5 Mining for Cancer Symptoms -- 10.6 Text Summarisation and Translation of Patient Record -- 10.6.1 Summarising the Patient Record -- 10.6.2 Other Approaches in Summarising the Patient Record -- 10.6.3 Summarising Medical Scientific Text -- 10.6.4 Simplification of the Patient Record for Laypeople -- 10.7 ICD-10 Diagnosis Code Assignment and Validation -- 10.7.1 Natural Language Generation from SNOMED CT -- 10.8 Search Cohort Selection and Similar Patient Cases -- 10.8.1 Comorbidities -- 10.8.2 Information Retrieval from Electronic Patient Records -- 10.8.3 Search Engine Solr -- 10.8.4 Supporting the Clinician in an Emergency Department with the Radiology Report -- 10.8.5 Incident Reporting -- 10.8.6 Hypothesis Generation -- 10.8.7 Practical Use of SNOMED CT -- 10.8.8 ICD-10 and SNOMED CT Code Mapping -- 10.8.9 Analysing the Patient's Speech -- 10.8.10 MYCIN and Clinical Decision Support -- 10.8.11 IBM Watson Health -- 10.9 Summary of Applications of Clinical Text Mining -- 11 Networks and Shared Tasks in Clinical Text Mining -- 11.1 Conferences, Workshops and Journals -- 11.2 Summary of Networks and Shared Tasks in Clinical Text Mining -- 12 Conclusions and Outlook -- 12.1 Outcomes -- References -- Index.