Evaluation Of Mechanism Of Insulin Measurement Based On Physiological,Biochemical And Multiparas

Diabetes has become one of the three major diseases that endanger human health,as have cardiovascular diseases and tumors,and the treatment of diabetics is becoming more and more important,according to the latest fluid data from the World Health Organization.Therefore,in the course of clinical treatment of diabetes,doctors need to always know the insulin level of diabetic patients,in order to grasp the clinical drug use and treatment effect,so the clinical insulin measurement and evaluation of diabetic patients is of great significanceIn this paper,the mechanism of insulin measurement in diabetic patients is studied in depth.Therefore,non-invisible,real-time feedback,convenient non-incasing blood glucose monitoring method is particularly important.the blood glucose value collected by the noninvasive glucose meter developed by our own research team and many other different physiological and biochemical data are used to establish an insulin secretion measurement model,and three different machine learning methods are used to realize the evaluation method of islet β cell function and insulin measurement prediction.The specific research work is as follows:(1)A similar number of type 2 diabetes patients and healthy people were selected,divided into two groups for comparative testing,to obtain their empty stomach blood sugar,insulin content,sugar tolerance and other physiological and bio-chemical parameters,the data obtained for pre-processing.(2)To plot the change curve of blood sugar and insulin concentration of two groups of study subjects at different time periods,compare and analyze the differences between the two from the line chart.(3)Build a machine learning model,the experimental data will be normalized after processing training,optimize parameters and adjust the structure of the model,so that it can get the best prediction results.(4)Using the completed machine learning model to predict the evaluation index of insulin,analyze and evaluate the function β islet cells.Through the experimental results,it can be found that the evaluation of insulin function based on machine learning model,the prediction accuracy is high,the fit advantage is above0.85,can evaluate insulin secretion function more accurately,can achieve effective monitoring,has good application value.In addition,the model algorithm based on extreme learning machine model and competitive neural network can provide new ideas for diabetes research and facilitate long-term tracking and identification.