Research On Disease-assisted Diagnosis Based On Artificial Intelligence And Fuzzy Mathematics

The development of computer technology and the rise of artificial intelligence provides new ideas for medical diagnosis.The neural network can change the weight of each index in the network achieving the processing of information,and effectively solve the Bottleneck problem of data processing in the medical diagnosis system.In addition,in the process of disease diagnosis,it is difficult to describe the concepts of “appetite loss” and “headache” with traditional mathematical methods.The emergence of fuzzy mathematics provides a good tool for solving such problems.It is possible to analyze the actual problems in medical diagnosis using the idea of fuzzy mathematics.The medical diagnosis system studied in this paper combines artificial intelligence with fuzzy mathematics to establish a diagnostic model.The model is designed to assist inexperienced physicians in the diagnosis of complex multi-index conditions.In this model,the BP neural network optimized by genetic algorithm is used to reduce the dimension of multi-indicator data,and then the related theories such as membership function,fuzzy measure and Choquet integral in fuzzy mathematics are used to calculate the index after dimension reduction.According to the calculation results and the experience of medical experts,the degree of illness of the examinee in the interval [0,1] is divided,comparing the calculated results with the standard values,the diagnostic results can be obtained.This model can simulate a doctor's diagnosis of the disease process and is very helpful for inexperienced doctors.The main contents of this paper are as follows:1.Using neural network and genetic algorithm to determine key indicators: using Matlab to realize simulation medical diagnosis system,self-learning adaptive data adjustment,obtaining key detection indicators needed for auxiliary diagnosis,the dimensionality of the test indicators is reduced when the correct rate of diagnosis is controllable.2.Establishing a medical aided diagnosis model using the relevant knowledge of fuzzy theory: introducing the concept of fuzzy measure,calculating the non-additive measure value under the interaction relationship of the test index;determining the membership function relationship between the test index and the disease;Calculating the integral value of the Choquet integral,Sugeno integral,and Wang integral of the membership function under non-additive measure,and the final disease diagnosis result is judged by the integral value.3.Design and implementation of disease-assisted diagnosis system: the system adopts MVC open source architecture,designs and develops the symptom indicators of the input examinee,and the system automatically calculates the membership degree of the disease to determine the important function module of the possibility of illness;add,delete,change,and check the maintenance of the examinee and medical data.