论文部分内容阅读
目的实现正常的窦性心律与心室纤颤信号的分类,从而检测出心室纤颤信号。方法该算法基于支持向量机技术,Hurst指数和相空间重构算法。待检测信号取自BIH-MIT和CU数据库,首先对待检测信号进行预处理,然后取滑动窗长度为3s计算出心电信号段的动力学指标值Hurst指数与相空间重构算法中的d值,最后把这两个参数作为特征向量输入到事先设计好的二分类支持向量机中,从而实现分类。结果成功实现了心室纤颤信号的分类,并通过计算该算法的灵敏度、特异性、预测性和准确度且与其他算法比较,可得新算法总体准确率优于其他算法。结论该算法可用于心电信号的检测,进行算法优化之后可嵌入到心电检测仪器中实现应用。
Objective To realize the classification of normal sinus rhythm and ventricular fibrillation signal and detect ventricular fibrillation signal. Methods The algorithm is based on support vector machine technology, Hurst exponent and phase space reconstruction algorithm. The signal to be detected is taken from the database of BIH-MIT and CU. The signal to be detected is preprocessed first, and then the dynamic index Hurst exponent of the electrocardiographic signal segment and the d value of the phase space reconstruction algorithm are calculated by taking the sliding window length as 3s Finally, these two parameters are input into the pre-designed two-class SVM as eigenvector to realize the classification. Results The classification of ventricular fibrillation signals was successfully achieved. Compared with other algorithms, the overall accuracy of the new algorithm was better than other algorithms by calculating the sensitivity, specificity, predictability and accuracy of the algorithm. Conclusion The algorithm can be used for the detection of ECG signals, which can be embedded in the ECG detection instrument after the algorithm is optimized.