【摘要】 提出了一种基于信号双谱分析的心室纤颤检测方法．首先，根据信号的准周期性，对信号进行了分段，并改进了信号三阶累积量的估计．然后，以特征值Ｐ表征信号，给出了心室纤颤信号的双谱检测算法，并结合其他算法，采取多特征值检测法进一步优化算法的性能．最后，在不同的噪声条件下，对不同的算法进行了比较，给出了微机实验的结果．实验表明：本文提出的心室纤颤检测算法达到了较优的灵敏度及误判率，并具有较强的抗噪声性能更多还原

【Abstract】 A novel algorithm based on bispectra analysis for detecting ventricular fibrillation(VF) was proposed. Firstly, according to the quasi period of the signal, the sequence was segmentalized and the estimation of cumulants was advanced. Secondly, the biomedical signal was detected by the characteristic parameter P . To improve analysis effect, two other parameters were also used in the algorithm. Finally, in computer simulation, three different algorithms were compared. The result shows that the proposed algorithm has high sensitivity and makes little mistake, and it was also robust when dealing with the noised signal.更多还原