论文部分内容阅读
目的 研究多因素个性化糖尿病决策支持系统 ,改善医疗效果 ,解决治疗中的个性化难题。方法 首次利用专家混合算法整合了普通药物、胰岛素、膳食、体能等多参数和多系数 ,为患者和专家提供合适的决策指导。描述了专家混合算法在糖尿病治疗中的应用实施步骤。建立了通用临床决策支持、糖尿病决策支持个性化糖尿病决策支持模型 ,用个性化模型对每天的治疗提供指导。结果 该系统的决策结果很接近训练数据 (r =0 .97± 0 .0 5 ,n =1 4 )和控制目标 (r=0 .95± 0 .0 6,n =1 4 ) .经验法指导治疗所得结果的标准差为 1 .81 ,个性化决策所得决策结果的标准差为 0 .2 1 .经验法与决策结果之间的标准差为 0 .783。结论 决策所得结果显示具有更好的均衡性 ,证明专家混合算法建立的个性化糖尿病决策支持系统是强有力的糖尿病决策工具 ,它可处理多系数和多参数决策数据。
Objective To study the multi-factor personalized diabetes decision support system to improve the medical effect and solve the individualized problems in treatment. Methods For the first time, we use expert hybrid algorithm to integrate multiple parameters such as common drugs, insulin, diet, physical fitness and multiple coefficients, and provide appropriate guidance to patients and experts. Describes the application of expert hybrid algorithm in the treatment of diabetes implementation steps. Established a universal clinical decision support, diabetes decision support personalized diabetes decision support model, with a personalized model for daily treatment to provide guidance. Results The decision-making results of this system are very close to the training data (r = 0 .97 ± 0. 05, n = 1 4) and control target (r = 0 .95 ± 0 .0 6, n = 1 4) The standard deviation of the results of the guideline treatment was 1.81, and the standard deviation of the decision-making results of the individualized decision-making was 0.21. The standard deviation of the empirical and decision-making results was 0.783. Conclusions The results obtained from the decision-making show a better balance. It is demonstrated that the personalized diabetes decision support system established by the expert hybrid algorithm is a powerful diabetes decision-making tool that can handle multi-coefficient and multi-parameter decision-making data.