基于支持向量机的室颤信号检测算法

目的:实现室颤信号与非室颤信号的分类,进而实现室颤信号的检测。方法:本文引入了一种基于支持向量机(Support Vec-tor Machine,SVM)和改进的越限区间算法(TCI)的新算法,其中支持向量机在处理分类和模式识别等问题中具有很大的优势。该算法采用4s的滑动窗技术,并利用改进后的越限区间算法(Threshold Crossing Interval,TCI)方法提取心电信号的特征。新算法的实现如下:在每一滑动窗内采用改进的后的绝对值阈值,计算中间2s内的平均越限间隔值。并以此TCI值作为特征参数,输入一个预先设计好的二分类支持向量机中,从而实现分类。结果:成功实现了室颤信号的检测,通过计算该方法的灵敏度、精确度、预测性和准确度且与其他方法相比较,可知此新算法总体可靠性优于其他方法。结论:该算法能够实现室颤信号的实时监测,且简单易行,易于实现,较适合实时的心电监测以及除颤仪器。

Ventricular Fibrillation Detection Algorithm Based on Support Vector Machine

Objective:To realize the discrimination of ventricular fibrillation(VF) and non-ventricular fibrillation(non-VF),accordingly detection of VF.Methods: The new algorithm was based on support vector machine(SVM) and the improved(TCI) algorithm.The SWM has great advantages in processing classification and pattern recognition.The new algorithm utilized 4-s-sliding-window technology and the improved TCI to extract features of ECG.It was implemented as follows: by using absolute thresholds,calculated average threshold crossing intervals of the middle 2s segment in every sliding window,and then input the TCI values into a binary classification support vector machine,finally,the discrimination was realized.Results: VF and non-VF were classified successfully.It shows that the new algorithm was superior to other classical algorithms by calculating quality parameters.Conclusions: This new algorithm can be used for real time VF detection.It is easier to implement and has greater advantages in real-time execution.It is suitable for ECG monitoring and defibrillator.