传统健康体检主要通过对单次体检指标进行横向比较和静态分析，忽略个体差异，纯粹从指标高低来判断体检者健康状况与身体状态，难以及时发现体检者可能存在的疾病隐患。由于个体体检指标具有样本量小、信息不确定、数据类型异构、影响因素构成复杂等特征，传统以大样本为基础的数学模型均难以适应此类小数据系统的建模要求。为此，通过建立适用于人体主要指标趋势预测的灰色系统模型（简称HIGM（1，1）模型），实现对人体主要健康指标的动态建模与趋势分析，从而可以掌握体检者未来一段时间身体指标的变化趋势及可能存在的疾病隐患。研究成果对提升体检效果、强化体检意义，促进灰色理论与现实问题的有效对接等方面，均具有积极意义。
Traditional health tests are often executed by horizontal comparisons and static analyses of a single physical examination result. The health condition and physical state are judged only based on the values of physical examination indicators, without the involvement of individual difference. Hence the potential disease cannot be discovered timely. Since the data of physical examination indicators from an individual are featured by small sample size, uncertain information, data isomerism and complicated influence factors, traditional mathematical models based on big data cannot be applied for such systems with a small data size. To this end, a novel grey system model, HIGM(1,1) for short, is established to forecast the development tendency of main human health indicators and then accomplish the dynamical modeling and tendency analysis of such human health indicators. Consequently, the changing tendency of physical indicators and potential disease can be obtained. The new HIGM(1,1) model is employed to forecast the index of serum creatinine which can reflect body kidney functions. The results show that the simulation and prediction errors of serum creatinine are less than those of the classical GM(1,1) model. The study findings have the important significance on promoting physical examination effect and intensifying physical examination meaning between grey theory and practical problems.
[1] 周伟杰, 张宏如, 党耀国, 等. 新息优先累加灰色离散模型的构建及应用[J]. 中国管理科学, 2017, 25(8): 140-148.
[2] 曾波, 刘思峰, 曲学鑫. 一种强兼容性的灰色通用预测模型及其性质研究[J]. 中国管理科学, 2017, 25(5): 150-156.
[3] 陆冬磊, 吴春峰, 段胜刚, 等. 应用灰色模型GM(1,1)预测上海市副溶血性弧菌引起的食源性疾病发病率[J]. 环境与职业医学, 2015, 32(8): 728-730.
[4] 蔡金钟. 疾病监测的灰色预测模型GM(1,1)研究[J].厦门大学学报(自然版),1995,(1): 121-126.
[5] Valipour M, Banihabib M E, Behbahani S M R. Comparison of the ARMA, ARIMA, and the autoregressive artificial neural network models in forecasting the monthly inflow of Dez dam reservoir[J]. Journal of Hydrology, 2013, 476:433-441.
[6] 华来庆, 申广荣, 熊林平, 等. ARIMA模型在黄瓜霜霉病疾病指数时间序列建模中的应用研究[J]. 第二军医大学学报, 2006, 27(7): 729-732.
[7] Krumm J. A markov model for driver turn prediction [J]. Sae World Congress, 2008, 22(1): 1-25.
[8] Dong M, He D. Hidden semi-Markov model-based methodology for multi-sensor equipment health diagnosis and prognosis[J]. European Journal of Operational Research, 2007, 178(3): 858-878.
[9] Simon Andrew Jones, Mark Patrick Joy, Jon Pearson. Forecasting demand of emergency care[J]. Health Care Management Science, 2002, 5(4): 297-305.
[10] 代小秋. 国内外症状监测系统发展现状及其评价框架[J]. 中华预防医学杂志, 2010, 44(6): 550-552.
[11] 王智, 刘薇, 兰燕. 基于BP神经网络的脑血管疾病院内感染风险预测模型研究[J]. 重庆医学, 2017, 46(12): 1691-1693.
[12] 樊楚, 贺向前, 于跃, 等. 基于数据挖掘技术建立的BP神经网络模型鉴别儿童川崎病与发热性疾病的研究[J]. 中国循证儿科杂志, 2017,(1): 22-26.
[13] 张文茜, 苏海霞, 尚磊, 等. 基于BP神经网络和RBF神经网络预测老年痴呆症疾病进展的对比研究[J]. 现代生物医学进展, 2017, 17(4): 738-741.
[14] Asiltürk ilhan, Cunkas M. Modeling and prediction of surface roughness in turning operations using artificial neural network and multiple regression method[M]. Oxford:Pergamon Press, Inc., 2011.
[15] Mcclelland G H, Irwin J R, Disatnik D, et al. Multicollinearity is a red herring in the search for moderator variables: A guide to interpreting moderated multiple regression models and a critique of Iacobucci, Schneider, Popovich, and Bakamitsos [J]. Behavior Research Methods, 2016, 49(1): 394-402.
[16] Shiota S, Okamoto Y, Okada G, et al. The neural correlates of the metacognitive function of other perspective: A multiple regression analysis study[J]. Neuroreport, 2017, 28(11): 671-676.
[17] Yoo H Y, Lee J H, Kim D S, et al. Enhancement of glucose yield from canola agricultural residue by alkali pretreatment based on multi-regression models[J]. Journal of industrial and engineering chemistry, 2017, 51: 303-311.
[18] 方益荣, 牛文柯, 卢巧玲, 等. 应用移动平均法预警学校结核病疫情的分析[J]. 疾病监测, 2017, 32(5): 418-422.
[19] Moscarelli M, Athanasiou T, Sevdalis N, et al. Controlled exponentially weighted moving average chart in cardiac surgery: A simulation study across 9 italian cardiac centers[J]. Seminars in Thoracic & Cardiovascular Surgery, 2016, 28(2): 253-258.
[20] 王永斌, 李向文, 柴峰, 等. 采用灰色-广义回归神经网络组合模型预测我国尘肺病发病人数的方法探讨[J]. 环境与职业医学, 2016, 33(10): 984-987+999.
[21] 江艺羡, 张岐山. 近似非齐次无偏GM(1,1)模型的递推解法及应用[J]. 控制与决策, 2015, 30(12): 2199-2204.
[22] 徐宁, 党耀国. 特征自适应型GM(1,1)模型及对中国交通污染排放量的预测建模[J]. 系统工程理论与实践, 2018, 38(1): 187-196.
[23] 杨保华, 赵金帅. 优化离散灰色幂模型及其应用[J]. 中国管理科学, 2016, 24(2): 162-168.
