Book cover: Clinical text mining
"Clinical Text Mining", new textbook written by Hercules Dalianis, Stockholm University.

The first book on clinical text mining is now published. The title of the book is "Clinical text mining: Secondary use of electronic patient records", published by Springer in year 2018, and is written by professor Hercules Dalianis. It is available as Open Access at https://www.springer.com

This book was written since there was a lack of a book - and specially a text book - describing the research area of clinical text mining, which means that from the large number of electronic medical records written today, extract new information, especially from the unstructured part of the patient record, the free text. A patient record is written by physicians and other healthcare professionals to describe symptoms, diagnosis and treatment of a patient, however, the record is rarely reused. There is very valuable information in a patient record, such as information on adverse drug effects, healthcare related infections, and other adverse events, but also early symptoms of cancer and comorbidities. To detect and predict these adverse effect and symptoms can improve healthcare for many patients. The methods for performing clinical text mining are mainly language technology and machine learning methods.

The book is based on more than 10 years of research, primarily on Swedish patient records work carried out by the Clinical Text Mining Group at DSV, as well as the work of hundreds of other international researchers. The book describes the history of the patient record from the ancient Egyptians to today's digital records that can be machine-readable. The book is also explaining the various medical classifications and terminologies such as ICD diagnostic codes, ATC drug codes, MeSH, UMLS and SNOMED CT. Furthermore, the book describes computational linguistics methods for processing clinical text that is difficult to process because it contains incomplete sentences, non-standard abbreviations and jargon, non-standard spellings of medical terms in Greek and Latin as well as non-standard abbreviations but also many misspellings. Part of the book describes the ethical and safety aspects of working with clinical text because it contains sensitive information about the patients that cannot be disclosed. The book finally describes a large number of applications in the area where electronic patient records are used as input data.