非高斯条件下基于粒子滤波的T波交替检测算法

针对动态心电图中非稳态、非线性和非高斯T波交替检测（T-wave alternans，TWA）的稳健检测问题，本文首先提出了一个TWA的非线性、拉普拉斯噪声的状态空间模型，然后采用粒子滤波方法对动态心电信号中的T波交替幅值进行估计，实现了心电信号中T波交替的定性和定量分析.仿真结果表明，在加入高斯噪声和非高斯噪声情况下，与经典的SM和MMA方法进行对比，本文提出的算法准确度和稳健性表现最好，并且本文算法能准确地实时地跟踪非稳态的TWA现象.

A T-wave Alternans Detection Algorithm Based on Particle Filtering in Non-Gaussian Environment

Many T-wave alternans (TWA) detecting algorithms have been proposed in the past,but most of them cannot provide an accurate and robust result,especially for non-stationary,non-linear and non-gaussian TWA in dynamic ECG.In this paper,a non-linear state space model with Laplace noise is proposed first,then particle filters is applied to estimate and analyze the T-wave alternans (TWA) amplitude in dynamic ECG.Simulated ECG signal with different TWA amplitudes and twadb database with clinical ECG records are evaluated by our proposed method.From the simulation results,our proposed algorithm is demonstrated the most accurate and robust for TWA detection when adding gaussian noise and non-gaussian noise,comparing to conventional spectral method (SM) and modified moving average (MMA) method.Besides,non-stationary TWA can be tracked accuratly in real-time by the proposed method.