The Research And Design Of Data Mining System For Brain Diseases

Since the 21 st century,with the major advancements in the application of computer technology in various industries,the application of computer technology to other industries has become a hot topic of research.The application in the medical field has achieved many results.After more than 30 years of practice and development,medical information processing has become an emerging edge discipline that integrates medical and computer technology and plays an important role in the development of medicine.The traditional medical field has accumulated vast amounts of medical data in the long development.How to effectively use the computer to process these data has become a hot research direction in the medical field.In recent years,technologies such as big data,cloud computing,and data mining have become increasingly mature,and research into traditional medical data has entered a new phase.The brain is the most important nerve center in humans.Electroencephalography(EEG)is a spontaneous,rhythmic electrical activity of the brain cell population recorded by the electrodes.In the clinical diagnosis and treatment of brain diseases,EEG has extremely important reference value.The research on EEG signal mining and analysis has been increasing.A large number of studies at home and abroad have shown that a lot of achievements have been made in the application of data mining methods and analysis of EEG signals.In this thesis,based on the brain care remote care system project,based on epileptic seizure EEG data from the EEG monitoring room of the Third People’s Hospital of Chengdu,a set of data mining system for brain diseases was studied and designed.It mainly includes the following aspects:First of all,it deeply studies the cross-industry data mining standards,analyzes the traditional data mining process and the research status of data mining in the medical field at home and abroad,and summarizes the excellent methods and theories.Secondly,the theories and methods of EEG signal feature extraction and selection at home and abroad are studied,and the methods for feature extraction and selection of EEG signals during epileptic seizures are described.The EEG data collected from hospitals during epileptic seizures are analyzed.The extraction of linear features and nonlinear features was performed using the ReliefF algorithm to process the characteristics of EEG signals,and experimental calculations and analyses were performed.Finally,it mainly completes the analysis of the needs of brain disease data mining system.The overall design of the system includes dividing the system into a data processing subsystem,an algorithm retrieval subsystem,a knowledge display subsystem,a user management subsystem,and the realization technology of the main functions of the system.,functional modules and main processes.The thesis is divided into five parts.The first part introduces the research significance,background and status quo of medical data mining,and the main work and organizational structure of this thesis.The second part studies and implements the feature extraction and feature selection algorithm of EEG signals.The third part makes a detailed analysis of the data mining system for brain diseases.The fourth part is the overall design of the brain disease data mining system and the design details of each module.The fifth part explains and displays the realization of the main functions of the system.At last,it summarizes and prospects the work of this thesis.