基于T波电交替检测的心源性猝死预警方法研究

Because of the sudden onset and rapid progress, it is difficult to treat the sudden cardiac death in time, which results in high mortality rate. Since China is in a high incidence of sudden cardiac death and the number of death ranks first in the world every year, its early detection and early diagnosis become one of the biggest challenges facing humanity and medicine in this century. At present, T-wave alternans (TWA) has been recognized as an independent, noninvasive and statistically significant core technology of effective prediction of sudden cardiac death. Then, this project will combine the electrophysiological mechanism and dynamic characteristics of human cardiac electrical activity to carry out the following investigations on the detection of T-wave alternans and the early warning of sudden cardiac death from the viewpoint of fault detection: the modeling, identification and feature extraction of the electrical activity system of the heart dynamics based on the advanced filtering algorithm and the learning theory, the dynamic mechanism modeling and analyzing of the abnormal electric activity based on bifurcation and structural stability theory, the early warning of sudden cardiac death and detecting of T-wave alternans under the mechanism of dynamical pattern recognition, and the simulation verification based on virtual heart model and the clinical data. The key to the project is to extract the intrinsic dynamics of the cardiac electrical activity process, establish the quantitative relationship between the dynamic information and the stability of the electrical activity process, and quickly and accurately detect the T wave alternans. The objective of this study is to establish a new idea for the early diagnosis of sudden death and to provide a basis for the diagnosis of sudden death by investigating the mechanism of sudden cardiac death and the establishment of a new method of T-wave alternating detection.

心源性猝死因发病突然、进展迅速而难以及时治疗，且死亡率很高，我国是心源性猝死高发地区，每年致死人数高居全球各国之首；因而如何早发现、早预警和早诊断已成为当前人类与医学领域所面临的最大挑战之一。而采用T波电交替则是目前国际公认的能有效预测心源性猝死的独立、无创且具有统计意义的核心技术。为此，本项目将结合人体心脏电活动的电生理和动力学本质特性，从故障诊断角度，开展针对T波电交替检测及心源性猝死预警问题的如下研究：基于先进滤波和学习理论的系统电活动动力学建模、辨识及特征提取，基于分岔和结构稳定理论的异常电活动过程的动力学机理建模，动态模式识别机制下的T波电交替检测和心源性猝死预警，针对虚拟心脏模型和临床数据的仿真验证。重点解决心脏电活动系统内在动力学特性的提取，动力学特征与系统稳定性量化关系的建立，快速准确的T波电交替检测等关键问题。研究成果的应用可为猝死的早期诊断提供新思路和临床辅助诊断依据。

心源性猝死因发病突然、进展迅速而难以及时治疗，且死亡率高。我国作为心源性猝死高发地带，每年猝死人数高居全球各国之首。近年来的研究和临床实验表明，猝死死亡率高居不下的主要原因是缺乏有效的早期诊断或预警技术，这一技术也是当前人类与医学领域所面临的最大挑战之一。T波电交替作为一种异常的心脏电信号，被国际公认为是当前预测恶性心律失常和心源性猝死的独立、无创且具有统计意义的最有价值的电生理检测指标。本项目针对心源性猝死预警问题，从心脏电活动的电生理和动力学本质特性入手，以T波电交替现象为对象，主要开展以下研究：1、基于动力学理论，研究离子机制下T波电交替现象的动力学改变过程，确定了T波电交替的动力学机理。选择HR模型为基础模型来仿真心肌细胞的电活动过程，以外界刺激电流以及反应细胞内Ca2+离子、K+离子浓度的参数为参变量，深入探究了不同参数条件下心脏电活动节律从正常到异常的发展过程，验证了医学上的T波电交替现象就是非线性动力学上的一种倍周期分岔行为。2、基于上述电活动过程分析结果，开展了心脏电活动过程的动力学建模和动态模式识别研究。利用不同参数条件下的动态模式构建动态模式库，基于结构稳定和确定学习理论，实现未知动态环境下电活动过程的动态识别，实现了倍周期分岔的提前检测。3、基于临床及官方公开的TWA数据，进一步验证上述T波电交替检测算法的有效性，结合模型和数据的验证结果，最终实现心源性猝死的预警。过程中，引入滤波算法实现心电数据的去噪，在确定学习理论框架下，基于RBF神经网络对正常和交替状态下的心电数据模式进行了动态表达，结合结构稳定和倍周期分岔理论，完成对待测心电模式进行动态识别，实现了T波电交替现象的提前检测。本项目的工作从机理层面建立起了医学与工科的桥梁，同时也基于动力学机理实现了心源性猝死的预警，有待进一步从软件层面系统化，拟为临床上心源性猝死的早诊断提供着实有效的技术辅助。