Predicting the risk of infrared thermal imaging technology on metabolic-related fatty liver disease based on random forest model
Objective:To explore the predictive effect of infrared thermal imaging on metabolism-related fatty liver disease using a random forest model.Methods:A total of 430 people who underwent physical examination in the physical examination center of our hospital from January 2021 to December 2022 were selected retrospectively.According to the results of the abdominal ultrasound examination,the patients were divided into fatty liver and control groups.Based on the structural theory of western medicine and the theory of traditional Chinese medicine,the positioning method was established,the infrared body region of traditional Chinese medicine was delineated and named,and the calorific value in the measured area was quantitatively analyzed.Logistic regression and random forest were used to construct the model of infrared thermal imaging to predict metabolism-related fatty liver disease.Results:The age,sex,governor pulse,eyes,double axils,double sides,upper focus,middle coke,lower coke,axillary-governor,axillary-middle coke,governor-eye,middle focus-lower focus were taken as independent variables,while fatty liver was taken as dependent variables.The main results are as follows:the total prediction accuracy of the model established by binary logistic regression is96.5%;the total prediction accuracy obtained by using the random forest model on the test set is 97.4%.Conclusion:The random forest model can accurately predict the risk of MAFLD according to infrared thermal imaging technology,which is helpful in improving the efficiency of early detection and diagnosis of MAFLD.
万方数据
维普
