Data Mining in Neurology
Antonio Candelieri, Giuliano Dolce, Francesco Riganello, Grey Walter · InTech eBooks · 2011
Data Mining intersects database technology, modelling techniques, statistical analysis, pattern recognition, and machine learning.It makes use of advanced tools for large databases management and automatic/semiautomatic analyses in order to identify significant trends and associations deemed informative because novel, implicit to the data, and of potential support in prediction and decision making.Methodological relevance and application in healthcare and biomedicine are increasing, with implications in fields as different as information management in healthcare organisation, public health, epidemiology, patient monitoring and management, signals and images analyses.It essentially represents an effective and efficient solution providing new predictive criteria for early diagnosis and prognosis, or supporting medical staffs in patient management such as in therapy planning and personalization.Knowledge extracted from pertinent clinical databases through data mining techniques may be new or suitable of integration with consolidated knowledge and improve reliability while reducing subjectivity in decision making processes.In this chapter we discuss about the general rationale underlying Data Mining and its peculiarities of application in the medical field, notably in the neurological domain.Relevant decision making problems, proposed solutions and open issues are summarized and the state-of-the-art of Data Mining in medicine and neurology is discussed in perspective.This review cannot and is not meant to be exhaustive, but should outline the potential use of Data Mining for supporting clinicians in their decision making. Rationale and backgroundData Mining was introduced in 1989 by Fayaad as a non-trivial process to identify reliable, novel, and potentially useful patterns in large data sets (Fayaad, 1996) though an iterative and multidisciplinary approach based on interaction with the application domain expert, data pre-processing, acquisition of consolidated knowledge, selection and use of the most suitable Data Mining methods, and evaluation and post-processing of the results.In this www.intechopen.