Medical Data Mining Application Based On Intelligent Algorithms

Intelligent algorithm is a kind of bionic algorithm inspired by natural phenomena or the mechanism of organism.With the rapid development of artificial intelligence,many algorithms,which applications become more and more widely,come into the world.This paper attempts to introduce intelligent algorithms into the research of stroke TCM Syndrome diagnosis Scale Construction.There are two important factors in Syndrome diagnosis Scale:one is Syndrome,another is four examinations.Syndrome refers to the summarization of human etiology and pathological changing in a certain stage of the disease development.Four examinations are the symptoms in four areas including human body information obtained by looking,hearing,smelling,questioning and touching,such as cough,convulsion,headache,and so on.Syndrome differentiation is a process that determines the nature as a certain "syndrome" and reveals the disease essence by analyzing and integrating all the information.Syndrome diagnosis Scale which services as a judge of "syndrome" is a significant ground for differentiating syndrome.For TCM Syndrome diagnosis Scale Construction,the main function of the intelligent algorithm is to optimize the weight for Mechanisms and Models of our research.Medical data mining based on intelligent algorithm is to dig out objective and accurate diagnostic criteria from massive medical data.This thesis is based on the Major State Basic Research Development Program ("973 Project") ---- "The Study on the Evaluation Criteria of the Diagnosis and Therapeutic Effects with the integration of Disease and Syndrome of Ischemic Stroke". Under the ground of Syndrome diagnosis Scale for stroke,data mining method is used to dig out an intuitive diagnostic criterion for stroke according to the large sample of forward-looking information consultation survey and research data.The essence of Syndrome diagnosis Scale for stroke is to dig out a mathematical model for stroke diagnosis,in which the main work is to find the dominant information about stroke acquired from four examinations.Weave this scale is to assign weights for the above information items,and then diagnose the "syndrome" according to the weighted sums.The main work for this stage is to find the decisive items and assign weights for them.Decisive items have been dug out in previous work,so this research focuses on assign weights for decisive items by digging weights from empirical data using intelligent algorithm and data mining theory.The main research contents are as follows:(1) Analyze the application prospect of this research and confirm the research target,then propose the main problem we would solve.After simply introducing the principle of Syndrome diagnosis Scale,we will discuss the key points-weights assignment and optimization.(2) Assign and optimize the weights of Syndrome diagnosis Scale by genetic algorithm(GA),particle swarm optimization(PSO),immune algorithm(IA).The work mainly includes:compare the results of different algorithms;analyze each algorithm's character,and then make a new algorithms special for our work.