基于线性时不变系统的家庭冠心病仿真预测模型研究

项目名称： 基于线性时不变系统的家庭冠心病仿真预测模型研究

项目编号： No.61201357

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 电子学与信息系统

项目作者： 吕勇强

作者单位： 清华大学

项目金额： 25万元

中文摘要： 冠心病，即冠状动脉粥样硬化心脏病，已经成为"第一杀手"，是世界卫生组织和各国重点防治的疾病之一，行之有效的院外非临床预测、预防措施是目前的研究热点。传统的冠心病预测方法和模型基本都存在预测时效性差、精度低、普适性弱、不适合家庭个人长期监测预防使用的缺点。本项目从信息技术角度出发，以家庭冠心病预测为目标，提出基于灰箱方法的线性时不变系统冠心病模型，拟通过归纳冠心病各种风险因子在发病过程中的作用和内在的关联特征建立模型结构，通过基于离散化近似模型的系统辨识确定模型，结合有效的连续、离散模型计算机仿真算法，对比临床检验结果对模型进行修正迭代，在模型基本完善的基础上再施以模型降阶处理，以期简化模型状态空间，增加家庭使用的实时性和简便性。本项目将模型的计算机仿真预测和冠心病临床实验紧密结合互动，确保最终模型的临床权威性，为国家相关应用研究打下理论基础。

中文关键词： 生理过程建模；风险因子；线性时不变系统；冠脉造影评分；血管压力响应指数

英文摘要： Cronary Heart Disease (CHD), also called Cronary Artery Disease, is often caused by atherosclerosis, which has been regardered as "First Killer" and one of the most important preventive dieseases of WHO and all over the world. Research on efficient non-clinical defense methods are becoming the best hot issues. Traditional CHD prediction methods and models are basically less timely, inprecise, non-pervaisve and not fit for long-time family use.This proposal introduces a grey-box based CHD modele with a unique informatics perspective, which employs linear time invariant system (LTI) and aims at family use. The model structure is setup based on concluding the traditional varisous CHD risk factors and is identified via the method of discrete approximation. The model is verified through combined computer simulation and clinical testing. The simulation takes both continuous and discrete formulations and interacts with clinical experiments,which can finally promote the original model results.In the end, the model will also be simplified by model order reduction, which is specially considered for applicable family use.The research will keep close connections between computer simulation/prediction and clinical verification, which guarantees the real authority of the model and prepares for other related application resea

英文关键词： Physiological modeling；Risk factor；Linear time invariant system；SYNTAX；Stress-induced vascular response index (sVRI)