Study On Metabolic Syndrome Based On Risk Score

Metabolic syndrome refers to the pathological state of metabolic disorder of proteins,fats,carbohydrates and other substances in human body.It is a complex group of metabolic disorders,which can be manifested as different combinations of various diseases in different patients.In addition to genetic factors,as the unhealthy lifestyles of people become more and more common,the proportion of people suffering from metabolic syndrome increases.As a complex of multiple diseases,with hypertension,hyperglycemia,hyperlipidemia,hyperuricemia and a series of serious complications,the mortality rate of patients with metabolic syndrome is much higher than that of patients suffering from only one of the diseases,which makes it an urgent problem to accurately diagnose metabolic syndrome.At present,the diagnosis of metabolic syndrome is mainly based on several key metabolic indexes.It is unrealistic to obtain them just in one examination.In addition,there is also a certain error in judging only by the key metabolic indexes.Therefore,it is the primary problem to propose a diagnosis method of metabolic syndrome based on incomplete examination data.Risk scoring system is an evaluation system to diagnose the occurrence of diseases at the next moment through multiple physical examination indexes.Some of the physical examination index data are unavailable.In this dissertation,the risk scoring system is introduced into the evaluation process of metabolic syndrome.And the risk of metabolic syndrome is diagnosed based on the incomplete physical examination index data of patients.More than 8,000 samples are collected from 2018 to 2021.And the samples are filtered and preprocessed to build the sample dataset.Through different patterns,the weight of indexes in risk score are defined.For the changing patterns and the threshold that are not in literatures,Particle Swarm Optimization is applied.And the K-fold of the data is applied to implement the cross-validation.After the changing patterns are determined,the weight of sampling index is determined by entropy method.Then the risk score is obtained.Based on the threshold,the sample is judged as metabolic syndrome.After completing the construction of the risk scoring system,this dissertation discusses the gender bias in the risk scoring system through the deciles and percentiles of risk scores and indexes.Objective To further refine the risk scoring system of metabolic syndrome according to different gender,so as to realize accurate judgment.The risk scoring system of metabolic syndrome constructed in this dissertation provides a new method and reliable theoretical basis for the early diagnosis of metabolic syndrome.Compared with traditional diagnosis methods,the diagnosis system based on risk score can reduce the cost of patients,and has the characteristics of low cost,convenience,rapidity and effectiveness.For patients,providing opportunities for early detection and treatment can reduce or eliminate the influence of complications of metabolic syndrome on people’s quality of life,and play a promoting role in improving people’s health.