Data Mining in Medical and Biological Research

Abstract

Data mining is the research area involving powerful processes and tools that allow an effective analysis and exploration of usually large amounts of data. In particular, data mining techniques have found application in numerous different scientific fields with the aim of discovering previously unknown patterns and correlations, as well as predicting trends and behaviors.In the meantime, during the current century of biomedical sciences, biology and medicine have undergone tremendous advances in their technologies and therefore have generated huge amounts of biomedical information. This book intends to bring together the most recent advances and applications of data mining research in the promising areas of medicine and biology from around the world. Seventeen chapters, twelve related to medical research and five focused on the biological domain, describe interesting applications, motivating progress and worthwhile results.Chapter 1 presents a novel concept of improving the classification accuracy for common learning algorithms by uncovering the potential relevance of features through multivariate interdependent attribute preprocessing methods.Chapter 2 introduces a hybrid approach, combining discrete wavelet transform and neural networks, for the classification of a complex dataset such as an electroencephalography time series and claims that a high degree of classification accuracy can be achieved.Chapter 3 explores the possibility that DNA viruses play an important role in development of breast tumors, using artificial neural networks and agglomerative hierarchical clustering techniques.Chapter 4 investigates the research hypothesis of developing certain domain ontology from discovered rules. To practically examine this assumption, the proposed approach integrates different data mining techniques to assist in developing a set of representative consensual concepts of the underlying domain.Chapter 5 describes the process of medical knowledge acquisition and outlines the requirements for producing knowledge that can be trusted and applied in a clinical setting. This work demonstrates that the individual needs of medical professionals can be addressed and that data mining can be a valuable tool in the diagnostic and decision making toolkit.Chapter 6 provides an overview of the various technologies in a radiology department that allow data mining, and describes case-examples utilizing data mining technologies.Chapter 7 demonstrates the results of a nationwide study that applied data and text mining techniques to medical near-miss case data related to medicines and medical equipments. This study also analyzed the causal relationship of occurrences of the near-miss cases and the opinions on the counter measures against them. VI Chapter 8 presents a complex clinical problem that has been addressed using particle swarm optimization, a suitable approach for data mining subject to a rule base which defines the quality of rules and constancy with previous observations. Chapter 9 focuses on the description of a model for mo...

Adaptive Probabilistic Neuro-Fuzzy System and its Hybrid Learning in Medical Diagnostics Task

“…When the first vector from the training dataset is fed to the system input x(1)=(x1(1),...,xi(1),...,xn( 1))−1 (it does not matter which of the classes Clj it belongs to), the center-"winner" wli*( 0) is determined at the beginning, which is the nearest xi (1) in the sense of distance, i.e., (13) (14) After this center "winner" is pulled up to the input signal xi(1) component according to the expression:…”