基于武汉市某社区居民电子健康档案的高血压预测模型  

Prediction model of hypertension based on electronic health records of a community resident in Wuhan

作　　者：万琦 王威[2] 黄薇[3] 容芷君[2] WAN Qi, WANG Wei, HUANG Wei, RONG Zhi-jun(Wuhan Children＇s Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei 430015, Chin)

机构地区：[1]华中科技大学同济医学院附属武汉儿童医院中西医结合科,湖北武汉430015 [2]武汉科技大学工业工程系,湖北武汉430081 [3]武汉市医学科学研究所,湖北武汉430014

出　　处：《现代预防医学》2018年第6期1030-1033,共4页Modern Preventive Medicine

摘　　要：目的分析武汉市某社区居民的健康档案数据,研究高血压患病的危险因素并建立基于神经网络的高血压预测模型,对血压值进行预测,为高血压预防提供依据并探索医疗健康大数据的应用模式。方法收集数据并进行处理,结合χ2检验与多因素Logistic回归模型分析高血压的危险因素。将危险因素作为输入,血压值作为输出建立BP神经网络预测模型。结果统计分析显示年龄、BMI、空腹血糖、体育锻炼、饮食习惯、心理状况为高血压患病的影响因素。神经网络预测结果显示收缩压与舒张压预测值的平均相对误差分别为9.25%和9.02%。结论 BP神经网络模型的预测准确度较好,能够对血压值进行预测,为高血压的预测与医疗健康大数据的应用提供了一种方法。Objective To survey the risk factors of hypertension and establish a neural network prediction model for hypertension by analyzing the health data of residents of a community in Wuhan. To provide the evidence for the prevention of hypertension and explore the application of medical health data. Methods The data were collected and analyzed, the risk factors of hypertension were analyzed by Chi-square test and multivariate logistic regression model. The BP neural network prediction model was established by using the risk factors as input and blood pressure as output. Results Statistical analysis showed that age, BMI, fasting blood-glucose, physical exercise, eating habits and psychological status were the influencing factors of hypertension. Neural network predictions showed that mean relative error of systolic and diastolic blood pressure predicted value were 9.25% and 9.02%, respectively. Conclusion The predictive accuracy of BP neural network model is acceptable, so BP neural network model can be used to predict blood pressure values, and can provides a method for the prediction of hypertension and the application of medical big data.

关 键 词：高血压 BP神经网络 预测模型 

分 类 号：R544.1[医药卫生—心血管疾病]