Evaluation Of Human Insulin Secretion Level And Non-invasive Blood Glucose Measurement

According to a survey by the World Health Organization,diabetes has become the third leading killer after cardiovascular and tumor diseases.Diabetes is mainly diagnosed through invasive testing.The testing process is likely to bring physical and psychological pain to patients,and wounds will be infected due to inadequate protective measures.Non-invasive blood glucose measurement technology has become a hot research topic in recent years because of its non-invasive,non-infectious and real-time monitoring characteristics.The blood glucose meter developed based on the energy metabolism conservation method can be used for the detection and identification of diabetes.This type of blood glucose meter has the characteristics of simple operation and non-invasive,and can realize the function of real-time monitoring of the blood glucose level of people with diabetes.This paper studies the evaluation and prediction of insulin,and explores the influence of the mathematical model of conservation of energy metabolism in noninvasive blood glucose detection.A method based on artificial neural network was proposed to study insulin evaluation and prediction models,and the prediction effects obtained by several artificial neural networks were compared.The specific work was as follows.Firstly,take type 2 diabetes patients and normal glucose tolerance groups as the research objects,organize them to conduct oral glucose tolerance tests,obtain multiple physiological parameters including glycated hemoglobin,blood glucose and insulin,and process the obtained data accordingly;Secondly,the blood glucose level and insulin level curve were drawn,and the difference between the curves of patients with type 2 diabetes and normal glucose tolerance was further analyzed;Thirdly,build an artificial neural network model,input the physiological parameters of the sample into the model for training,and constantly adjust the model parameters to obtain the best prediction effect;Finally,the built model was used to predict islet cell function index and insulin resistance index.The experimental results show that the predicted value of the insulin evaluation index and the true value obtained by the human insulin evaluation model based on artificial neural network had a good correlation,and the correlation of the islet cell function index reaches nearly 90%,indicating that the model can better Evaluation of insulin secretion.Further,according to the action of insulin,a correction method is proposed for the noninvasive blood glucose detection algorithm model based on the energy metabolism conservation method,which provides a new idea for its optimization.